------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication\re
> sults_coercive_imbalance.log
  log type:  text
 opened on:  17 Aug 2023, 12:24:14

. 
. 
. *###############################################################################
. *       1 DATA PREPARATION DESCRIBED IN TEXT
. *###############################################################################
. 
. *Exclude great powers (cf. Section 5.1)
. drop if country=="France" | country=="Germany" | country=="Russian Federation" | country
> =="United Kingdom" | country=="United States" 
(965 observations deleted)

. 
. *Exlude country-years with less than 1 million (cf. Section 5)
. drop if tpop<1000
(1,624 observations deleted)

. 
. *Data in four-year intervals (cf. Section 5)
. drop if year/4!=round(year/4)
(22,983 observations deleted)

. drop if year<1850 //this limits secures that we have variation in mtechiv_exf in the ent
> ire sample
(1,329 observations deleted)

. sort country year

. egen yearno=group(year)

. xtset ccode yearno

Panel variable: ccode (unbalanced)
 Time variable: yearno, 1 to 41, but with gaps
         Delta: 1 unit

. 
. *truncate (cf. Section 5)
. replace lagwe_lngdpc=min(lagwe_lngdpc,100)  if lagwe_lngdpc!=.
(1,716 real changes made)

. replace lagwe_mtech_exf=min(lagwe_mtech_exf,100) if lagwe_mtech_exf!=.
(329 real changes made)

. replace lagwe_mtech_exf_iv_n2=min(lagwe_mtech_exf_iv_n2,100) if lagwe_mtech_exf_iv_n2!=.
(0 real changes made)

. replace lagwe_mtech_s=min(lagwe_mtech_s,100)
(3,609 real changes made)

. 
. *generating variable to keep sample constant
. gen balance=1 if L.lagwe_mtech_exf!=. & L.lagwe_mtech_exf_iv_n2!=. & L.lagwe_lngdpc!=.
(4,110 missing values generated)

. label variable balance "Balance"

. 
. *###############################################################################
. *       2 ALL REGRESSION TABLES IN THE MAIN TEXT
. *###############################################################################
. 
. //Table 2: Arms Technology and Democracy
. reghdfe v2x_polyarchy L.lagwe_mtech_exf L.lagwe_mtech_s L.lagwe_lngdpc  if balance==1, v
> ce(cluster country) absorb(year country) 
(dropped 3 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      2,368
Absorbing 2 HDFE groups                           F(   3,    142) =       5.19
Statistics robust to heteroskedasticity           Prob > F        =     0.0020
                                                  R-squared       =     0.7823
                                                  Adj R-squared   =     0.7639
                                                  Within R-sq.    =     0.0309
Number of clusters (country) =        143         Root MSE        =    13.3620

                                 (Std. err. adjusted for 143 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .1264733   .0519344     2.44   0.016     .0238089    .2291377
                |
  lagwe_mtech_s |
            L1. |   .0276876   .0330348     0.84   0.403    -.0376159    .0929912
                |
   lagwe_lngdpc |
            L1. |  -.1173293   .0388635    -3.02   0.003    -.1941551   -.0405035
                |
          _cons |   41.70162   2.360128    17.67   0.000     37.03609    46.36714
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
        year |        40           0          40     |
     country |       143         143           0    *|
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table2.tex", keep(L.lagwe_mtech_exf L.lagwe_mte
> ch_s L.lagwe_lngdpc) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 addstat(Co
> untries, e(df_a_redundant)) replace sortvar(L.lagwe_mtech_exf L.lagwe_mtech_s  L.lagwe_l
> ngdpc) cttop(OLS)
coercive_imbalance_tables/Table2.tex
dir : seeout

. reghdfe v2x_polyarchy L.lagwe_mtech_exf L.lagwe_mtech_exf_weu L.lagwe_lngdpc  if balance
> ==1, vce(cluster country) absorb(country i.weu#i.year year)
(dropped 3 singleton observations)
(MWFE estimator converged in 10 iterations)

HDFE Linear regression                            Number of obs   =      2,368
Absorbing 3 HDFE groups                           F(   3,    142) =      12.63
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.8040
                                                  Adj R-squared   =     0.7836
                                                  Within R-sq.    =     0.0561
Number of clusters (country) =        143         Root MSE        =    12.7932

                                     (Std. err. adjusted for 143 clusters in country)
-------------------------------------------------------------------------------------
                    |               Robust
      v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------------+----------------------------------------------------------------
    lagwe_mtech_exf |
                L1. |    .192811   .0518063     3.72   0.000     .0903997    .2952223
                    |
lagwe_mtech_exf_weu |
                L1. |  -.2646104    .243448    -1.09   0.279    -.7458611    .2166402
                    |
       lagwe_lngdpc |
                L1. |  -.1583376   .0363963    -4.35   0.000    -.2302863    -.086389
                    |
              _cons |   43.25142    2.34412    18.45   0.000     38.61754     47.8853
-------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
    weu#year |        79           0          79     |
        year |        40          40           0     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table2.tex", keep(L.lagwe_mtech_exf L.lagwe_mte
> ch_exf_weu L.lagwe_lngdpc) label  addtext(Sample, Full) bdec(2) dec(2) nocons nor2 addst
> at(Countries, e(df_a_nested))  sortvar(L.lagwe_mtech_exf  L.lagwe_mtech_s L.lagwe_mtech_
> exf_weu L.lagwe_lngdpc) cttop(OLS) append 
coercive_imbalance_tables/Table2.tex
dir : seeout

. reghdfe v2x_polyarchy L.lagwe_mtech_exf L.lagwe_lngdpc  if balance==1 & weu==0, vce(clus
> ter country) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,882
Absorbing 2 HDFE groups                           F(   2,    126) =      15.50
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7461
                                                  Adj R-squared   =     0.7212
                                                  Within R-sq.    =     0.0636
Number of clusters (country) =        127         Root MSE        =    12.4616

                                 (Std. err. adjusted for 127 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .1956263   .0509436     3.84   0.000     .0948105    .2964421
                |
   lagwe_lngdpc |
            L1. |  -.1692529   .0466018    -3.63   0.000    -.2614766   -.0770293
                |
          _cons |   39.12373   3.553731    11.01   0.000     32.09101    46.15646
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table2.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Countries, e(df
> _a_redundant)) cttop(OLS) append
coercive_imbalance_tables/Table2.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc  if
>  balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    10.21
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =  -0.0102
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0102
Residual SS             =  286992.0295                Root MSE      =    12.49

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5487931   .1751893     3.13   0.002     .2020988    .8954875
                |
   lagwe_lngdpc |
            L1. |  -.2003625   .0505487    -3.96   0.000     -.300397   -.1003281
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.932
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.378
                         (Kleibergen-Paap rk Wald F statistic):         39.846
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table2.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap
>  F-statistic, e( rkf), Countries, e(df_a_redundant)) cttop(2SLS)   append
coercive_imbalance_tables/Table2.tex
dir : seeout

. xtabond2 v2x_polyarchy  L.(v2x_polyarchy lagwe_mtech_exf lagwe_lngdpc) i.year if balance
> ==1 & weu==0,  gmm(L.(v2x_polyarchy lagwe_mtech_exf lagwe_lngdpc), lag(1 2) collapse) iv
> (i.year) twostep robust   cluster(country)
Favoring speed over space. To switch, type or click on mata: mata set matafavor space, per
> m.
1852b.year dropped due to collinearity
2004.year dropped due to collinearity
Warning: Two-step estimated covariance matrix of moments is singular.
  Using a generalized inverse to calculate optimal weighting matrix for two-step estimatio
> n.
  Difference-in-Sargan/Hansen statistics may be negative.

Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: ccode                           Number of obs      =      1898
Time variable : yearno                          Number of groups   =       153
Number of instruments = 49                      Obs per group: min =         1
Wald chi2(42) = 145175.31                                      avg =     12.41
Prob > chi2   =     0.000                                      max =        40
                                   (Std. err. adjusted for clustering on country)
---------------------------------------------------------------------------------
                |              Corrected
  v2x_polyarchy | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
----------------+----------------------------------------------------------------
  v2x_polyarchy |
            L1. |   .7756187   .0398745    19.45   0.000     .6974661    .8537713
                |
lagwe_mtech_exf |
            L1. |   .0792653    .037419     2.12   0.034     .0059254    .1526051
                |
   lagwe_lngdpc |
            L1. |  -.0981791   .0299011    -3.28   0.001    -.1567842   -.0395741
                |
           year |
          1856  |  -4.219079   1.921987    -2.20   0.028    -7.986104   -.4520532
          1860  |  -5.902709   2.089897    -2.82   0.005    -9.998831   -1.806587
          1864  |  -4.399841   2.188549    -2.01   0.044    -8.689319    -.110364
          1868  |  -5.410787   1.902027    -2.84   0.004    -9.138692   -1.682883
          1872  |  -4.820562   2.032908    -2.37   0.018    -8.804989   -.8361345
          1876  |  -3.761954   1.935613    -1.94   0.052    -7.555686    .0317785
          1880  |  -5.682711   2.096361    -2.71   0.007    -9.791504   -1.573918
          1884  |  -5.093346   1.849266    -2.75   0.006     -8.71784   -1.468851
          1888  |  -4.954976   1.903658    -2.60   0.009    -8.686078   -1.223874
          1892  |  -4.896591   2.026817    -2.42   0.016    -8.869079   -.9241041
          1896  |   -4.14466   1.958827    -2.12   0.034     -7.98389    -.305431
          1900  |  -5.220893   1.860613    -2.81   0.005    -8.867627   -1.574158
          1904  |  -5.373461   1.779377    -3.02   0.003    -8.860976   -1.885946
          1908  |   -4.86153   1.744232    -2.79   0.005    -8.280161   -1.442899
          1912  |  -2.244334   1.723132    -1.30   0.193     -5.62161    1.132942
          1916  |  -5.362665   1.631594    -3.29   0.001    -8.560531   -2.164799
          1920  |    -2.6465   1.724169    -1.53   0.125    -6.025809    .7328097
          1924  |  -3.694918   1.553645    -2.38   0.017    -6.740006   -.6498294
          1928  |  -6.122078   2.206766    -2.77   0.006    -10.44726   -1.796897
          1932  |  -6.490682   2.119049    -3.06   0.002    -10.64394   -2.337421
          1936  |  -5.860755   1.990102    -2.94   0.003    -9.761282   -1.960227
          1940  |  -3.678288   2.446619    -1.50   0.133    -8.473573    1.116996
          1944  |  -5.793648   2.256084    -2.57   0.010    -10.21549   -1.371805
          1948  |   4.497354   2.904818     1.55   0.122    -1.195985    10.19069
          1952  |  -3.201129   2.052641    -1.56   0.119    -7.224231    .8219728
          1956  |  -3.559191   1.877034    -1.90   0.058    -7.238111    .1197284
          1960  |  -3.955563   1.858935    -2.13   0.033     -7.59901   -.3121171
          1964  |  -5.354004   1.270587    -4.21   0.000    -7.844309   -2.863699
          1968  |  -6.812808   1.496832    -4.55   0.000    -9.746544   -3.879071
          1972  |  -7.230687    1.30074    -5.56   0.000    -9.780091   -4.681283
          1976  |  -7.886338   1.372926    -5.74   0.000    -10.57722   -5.195452
          1980  |  -4.556259   1.492446    -3.05   0.002      -7.4814   -1.631117
          1984  |  -5.449014   1.396815    -3.90   0.000     -8.18672   -2.711308
          1988  |  -2.753641   1.619035    -1.70   0.089    -5.926892    .4196098
          1992  |   4.635284   1.851306     2.50   0.012      1.00679    8.263778
          1996  |   2.187857   .9710336     2.25   0.024     .2846661    4.091048
          2000  |  -.1447064   .9503727    -0.15   0.879    -2.007403     1.71799
          2008  |  -.7665775    .886464    -0.86   0.387    -2.504015      .97086
          2012  |   -.473818    .890559    -0.53   0.595    -2.219281    1.271646
                |
          _cons |   16.33344   2.740135     5.96   0.000     10.96288    21.70401
---------------------------------------------------------------------------------
Instruments for first differences equation
  Standard
    D.(1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year)
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(1/2).(L.v2x_polyarchy L.lagwe_mtech_exf L.lagwe_lngdpc) collapsed
Instruments for levels equation
  Standard
    1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year
    _cons
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    D.(L.v2x_polyarchy L.lagwe_mtech_exf L.lagwe_lngdpc) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -6.16  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =  -0.18  Pr > z =  0.855
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(6)    =   2.16  Prob > chi2 =  0.905
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(6)    =   2.55  Prob > chi2 =  0.863
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
    Hansen test excluding group:     chi2(3)    =   0.94  Prob > chi2 =  0.815
    Difference (null H = exogenous): chi2(3)    =   1.60  Prob > chi2 =  0.658


. outreg2 using "coercive_imbalance_tables/Table2.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc L.v2x_polyarchy) label addtext(Sample, Excl. WEU) bdec(2) dec(2)  nocons nor2 noni  
> addstat(Countries, e(N_g) ) cttop(System GMM) append
coercive_imbalance_tables/Table2.tex
dir : seeout

. 
. 
. //Table 3: Arms technology and Bureaucracy. Panel A
. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc  if 
> balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  2,   126) =     2.53
                                                      Prob > F      =   0.0835
Total (centered) SS     =  313810.0553                Centered R2   =  -0.0975
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.0975
Residual SS             =  344399.7399                Root MSE      =    13.71

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.5618462   .2645546    -2.12   0.036    -1.085392   -.0383005
                |
   lagwe_lngdpc |
            L1. |   .0838931   .0568984     1.47   0.143    -.0287072    .1964935
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.019
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.914
                         (Kleibergen-Paap rk Wald F statistic):         40.037
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3a.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Cou
> ntries, e(df_a_redundant)) groupvar("Panel_A") replace
coercive_imbalance_tables/Table3a.tex
dir : seeout

. ivreghdfe v2x_corr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc  if bal
> ance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1873
                                                      F(  2,   126) =     3.16
                                                      Prob > F      =   0.0457
Total (centered) SS     =  230424.5216                Centered R2   =  -0.1541
Total (uncentered) SS   =  230424.5216                Uncentered R2 =  -0.1541
Residual SS             =  265933.3917                Root MSE      =    12.05

---------------------------------------------------------------------------------
                |               Robust
       v2x_corr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.5517349   .2410453    -2.29   0.024    -1.028756   -.0747134
                |
   lagwe_lngdpc |
            L1. |   .0908237   .0566139     1.60   0.111    -.0212136     .202861
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.014
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.682
                         (Kleibergen-Paap rk Wald F statistic):         40.058
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3a.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Cou
> ntries, e(df_a_redundant))  groupvar("Panel_A") append
coercive_imbalance_tables/Table3a.tex
dir : seeout

. ivreghdfe v2x_execorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc  if 
> balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1884
                                                      F(  2,   126) =     4.97
                                                      Prob > F      =   0.0083
Total (centered) SS     =  354604.2258                Centered R2   =  -0.1816
Total (uncentered) SS   =  354604.2258                Uncentered R2 =  -0.1816
Residual SS             =  419005.0855                Root MSE      =    15.08

---------------------------------------------------------------------------------
                |               Robust
    v2x_execorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.7331007     .26642    -2.75   0.007    -1.260338   -.2058634
                |
   lagwe_lngdpc |
            L1. |   .1202181   .0607921     1.98   0.050    -.0000876    .2405238
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.899
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.842
                         (Kleibergen-Paap rk Wald F statistic):         39.903
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3a.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Cou
> ntries, e(df_a_redundant))  groupvar("Panel_A") append
coercive_imbalance_tables/Table3a.tex
dir : seeout

. ivreghdfe v2clrspct  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)   L.lagwe_lngdpc  if b
> alance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1877
                                                      F(  2,   126) =    11.17
                                                      Prob > F      =   0.0000
Total (centered) SS     =  9588823.878                Centered R2   =  -0.0052
Total (uncentered) SS   =  9588823.878                Uncentered R2 =  -0.0052
Residual SS             =  9638299.413                Root MSE      =    72.47

---------------------------------------------------------------------------------
                |               Robust
      v2clrspct | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   2.794225   .9936444     2.81   0.006     .8278314    4.760618
                |
   lagwe_lngdpc |
            L1. |  -.9947646   .2353921    -4.23   0.000    -1.460599   -.5289305
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.023
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.936
                         (Kleibergen-Paap rk Wald F statistic):         39.952
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3a.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Cou
> ntries, e(df_a_redundant))  groupvar("Panel_A") append
coercive_imbalance_tables/Table3a.tex
dir : seeout

. *Table 3: Arms technology and Bureaucracy. Panel B
. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc L.de
> mocracy  if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  3,   126) =     3.91
                                                      Prob > F      =   0.0104
Total (centered) SS     =  313810.0553                Centered R2   =  -0.0699
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.0699
Residual SS             =  335737.9158                Root MSE      =    13.54

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.5210742    .275085    -1.89   0.060    -1.065459     .023311
                |
   lagwe_lngdpc |
            L1. |   .0686974   .0573717     1.20   0.233    -.0448395    .1822344
                |
      democracy |
            L1. |  -4.214339   2.708006    -1.56   0.122    -9.573404    1.144726
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.683
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              207.015
                         (Kleibergen-Paap rk Wald F statistic):         38.664
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.democracy
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3b.tex",  keep(L.lagwe_mtech_exf) label bd
> ec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))groupvar("Panel_B") replace
coercive_imbalance_tables/Table3b.tex
dir : seeout

. ivreghdfe v2x_corr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc L.democ
> racy  if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1873
                                                      F(  3,   126) =     2.80
                                                      Prob > F      =   0.0427
Total (centered) SS     =  230424.5216                Centered R2   =  -0.1499
Total (uncentered) SS   =  230424.5216                Uncentered R2 =  -0.1499
Residual SS             =  264958.7953                Root MSE      =    12.03

---------------------------------------------------------------------------------
                |               Robust
       v2x_corr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   -.545926   .2564375    -2.13   0.035    -1.053408   -.0384437
                |
   lagwe_lngdpc |
            L1. |   .0886546    .055611     1.59   0.113     -.021398    .1987073
                |
      democracy |
            L1. |  -.6016264   2.707297    -0.22   0.824    -5.959288    4.756035
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.683
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              206.841
                         (Kleibergen-Paap rk Wald F statistic):         38.696
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.democracy
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3b.tex",  keep(L.lagwe_mtech_exf ) label b
> dec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_
> a_redundant)) groupvar("Panel_B") append
coercive_imbalance_tables/Table3b.tex
dir : seeout

. ivreghdfe v2x_execorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc L.de
> mocracy if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1884
                                                      F(  3,   126) =     4.46
                                                      Prob > F      =   0.0052
Total (centered) SS     =  354604.2258                Centered R2   =  -0.1719
Total (uncentered) SS   =  354604.2258                Uncentered R2 =  -0.1719
Residual SS             =  415577.8816                Root MSE      =    15.02

---------------------------------------------------------------------------------
                |               Robust
    v2x_execorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.7185143   .2836233    -2.53   0.013    -1.279797   -.1572322
                |
   lagwe_lngdpc |
            L1. |   .1149731   .0602376     1.91   0.059    -.0042352    .2341815
                |
      democracy |
            L1. |  -1.483243   3.011887    -0.49   0.623    -7.443679    4.477193
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.559
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              206.665
                         (Kleibergen-Paap rk Wald F statistic):         38.531
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.democracy
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3b.tex",  keep(L.lagwe_mtech_exf ) label b
> dec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_
> a_redundant)) groupvar("Panel_B") append
coercive_imbalance_tables/Table3b.tex
dir : seeout

. ivreghdfe v2clrspct  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)   L.lagwe_lngdpc L.dem
> ocracy if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1877
                                                      F(  3,   126) =    14.33
                                                      Prob > F      =   0.0000
Total (centered) SS     =  9588823.878                Centered R2   =   0.0482
Total (uncentered) SS   =  9588823.878                Uncentered R2 =   0.0482
Residual SS             =  9126766.573                Root MSE      =    70.54

---------------------------------------------------------------------------------
                |               Robust
      v2clrspct | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |    2.37406   1.012948     2.34   0.021     .3694655    4.378654
                |
   lagwe_lngdpc |
            L1. |  -.8401882   .2387354    -3.52   0.001    -1.312638   -.3677379
                |
      democracy |
            L1. |   43.02217   12.04596     3.57   0.001     19.18356    66.86078
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.668
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              205.958
                         (Kleibergen-Paap rk Wald F statistic):         38.535
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.democracy
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table3b.tex",  keep(L.lagwe_mtech_exf) label bd
> ec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))groupvar("Panel_B") append
coercive_imbalance_tables/Table3b.tex
dir : seeout

. 
. //Table 4: Robustness of 2SLS-results 
. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.n
> 1_democracy_avg L.n2_democracy_avg if balance==1 & weu==0, robust cluster(country) absor
> b(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  4,   126) =    12.38
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0366
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0366
Residual SS             =  273687.3575                Root MSE      =     12.2

----------------------------------------------------------------------------------
                 |               Robust
   v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |   .4668069   .1853742     2.52   0.013     .0999568    .8336569
                 |
    lagwe_lngdpc |
             L1. |  -.1987217   .0481884    -4.12   0.000     -.294085   -.1033584
                 |
n1_democracy_avg |
             L1. |   7.646955   3.623861     2.11   0.037       .47544    14.81847
                 |
n2_democracy_avg |
             L1. |   2.736598    5.34315     0.51   0.609    -7.837338    13.31053
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.398
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              191.250
                         (Kleibergen-Paap rk Wald F statistic):         37.092
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.n1_democracy_avg L.n2_democracy_avg
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addtext(ColonizerxTime-FE, ) addstat(Kleibergen-Pa
> ap F-statistic, e( rkf), Countries, e(df_a_redundant)) cttop(Spillovers)  replace  
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.n
> 1_interstatewar L.n2_interstatewar L.n1_civilwar L.n2_civilwar  if balance==1 & weu==0, 
> robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  6,   126) =     5.82
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0007
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0007
Residual SS             =  283904.1803                Root MSE      =    12.44

----------------------------------------------------------------------------------
                 |               Robust
   v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |   .5260894   .1758369     2.99   0.003     .1781132    .8740655
                 |
    lagwe_lngdpc |
             L1. |  -.2003176    .050039    -4.00   0.000    -.2993432   -.1012919
                 |
n1_interstatewar |
             L1. |   .7088044   1.255897     0.56   0.573    -1.776579    3.194188
                 |
n2_interstatewar |
             L1. |    1.22631   .9935615     1.23   0.219    -.7399188    3.192539
                 |
     n1_civilwar |
             L1. |   -.836317   .9408749    -0.89   0.376    -2.698281    1.025647
                 |
     n2_civilwar |
             L1. |  -.0694333   .9813033    -0.07   0.944    -2.011404    1.872537
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.432
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              210.293
                         (Kleibergen-Paap rk Wald F statistic):         37.503
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.n1_interstatewar L.n2_interstatewar
                      L.n1_civilwar L.n2_civilwar
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant)) cttop(Spillovers) append 
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.d
> istw1_democracy L.distw1_interstatewar L.distw1_civilwar L.distw1_lngdp  if balance==1 &
>  weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  6,   126) =     5.10
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =   0.0161
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0161
Residual SS             =  279528.7114                Root MSE      =    12.34

--------------------------------------------------------------------------------------
                     |               Robust
       v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
     lagwe_mtech_exf |
                 L1. |    .540632   .2023748     2.67   0.009     .1401382    .9411259
                     |
        lagwe_lngdpc |
                 L1. |  -.1717192   .0459544    -3.74   0.000    -.2626615   -.0807768
                     |
    distw1_democracy |
                 L1. |   4.421407    18.3755     0.24   0.810    -31.94317    40.78598
                     |
distw1_interstatewar |
                 L1. |   .3506999   8.533305     0.04   0.967    -16.53646    17.23786
                     |
     distw1_civilwar |
                 L1. |  -1.181058    14.7407    -0.08   0.936    -30.35246    27.99035
                     |
        distw1_lngdp |
                 L1. |  -24.28214   10.47803    -2.32   0.022    -45.01785   -3.546433
--------------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             20.339
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              151.070
                         (Kleibergen-Paap rk Wald F statistic):         28.752
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.distw1_democracy L.distw1_interstatewar
                      L.distw1_civilwar L.distw1_lngdp
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant)) cttop(Spillovers) append  
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.i
> nterstatewar_3y L.civilwar_3y  if balance==1 & weu==0, robust cluster(country) absorb(co
> untry year)
(dropped 18 singleton observations)
Warning: time variable yearno has 20 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1812
                                                      F(  4,   126) =     6.42
                                                      Prob > F      =   0.0001
Total (centered) SS     =  275213.9494                Centered R2   =  -0.0278
Total (uncentered) SS   =  275213.9494                Uncentered R2 =  -0.0278
Residual SS             =  282860.1368                Root MSE      =    12.65

----------------------------------------------------------------------------------
                 |               Robust
   v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |   .6270155   .1876416     3.34   0.001     .2556782    .9983527
                 |
    lagwe_lngdpc |
             L1. |  -.2067142   .0505232    -4.09   0.000    -.3066981   -.1067304
                 |
interstatewar_3y |
             L1. |  -1.511121   1.447874    -1.04   0.299     -4.37642    1.354178
                 |
     civilwar_3y |
             L1. |  -.3211159   1.307116    -0.25   0.806    -2.907859    2.265628
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.509
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              197.264
                         (Kleibergen-Paap rk Wald F statistic):         39.632
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.interstatewar_3y L.civilwar_3y
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant)) cttop(War & Alliances) append
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.a
> lly_USA L.ally_Russia if balance==1 & weu==0, robust cluster(country) absorb(country yea
> r)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  4,   126) =    12.11
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0578
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0578
Residual SS             =  267669.9545                Root MSE      =    12.07

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |    .481112    .159482     3.02   0.003     .1655017    .7967223
                |
   lagwe_lngdpc |
            L1. |  -.1957571   .0434456    -4.51   0.000    -.2817348   -.1097795
                |
       ally_USA |
            L1. |   12.89828   3.319328     3.89   0.000     6.329426    19.46713
                |
    ally_Russia |
            L1. |  -.7243786   2.092215    -0.35   0.730     -4.86481    3.416053
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.097
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              218.836
                         (Kleibergen-Paap rk Wald F statistic):         37.199
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.ally_USA L.ally_Russia
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant)) cttop(War & Alliances) append
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> ntrade    if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 18 singleton observations)
Warning: time variable yearno has 11 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     123                Number of obs =     1726
                                                      F(  3,   122) =     7.01
                                                      Prob > F      =   0.0002
Total (centered) SS     =  262723.6001                Centered R2   =  -0.0531
Total (uncentered) SS   =  262723.6001                Uncentered R2 =  -0.0531
Residual SS             =  276667.6807                Root MSE      =     12.8

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |    .701445   .2043462     3.43   0.001     .2969212    1.105969
                |
   lagwe_lngdpc |
            L1. |  -.1495033   .0468768    -3.19   0.002    -.2423006    -.056706
                |
        lntrade |
            L1. |   2.638049   1.141815     2.31   0.023     .3777121    4.898385
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.500
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              188.070
                         (Kleibergen-Paap rk Wald F statistic):         37.035
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lntrade
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       123         123           0    *|
        year |        35           0          35     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant)) cttop(Trade & Aid) append
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> naid    if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  3,   126) =    17.13
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =  -0.0020
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0020
Residual SS             =  284652.0618                Root MSE      =    12.44

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5394168   .1779868     3.03   0.003     .1871861    .8916475
                |
   lagwe_lngdpc |
            L1. |  -.2100204   .0454857    -4.62   0.000    -.3000352   -.1200056
                |
          lnaid |
            L1. |   .2159095   .2176896     0.99   0.323    -.2148918    .6467108
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.068
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              219.402
                         (Kleibergen-Paap rk Wald F statistic):         39.804
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lnaid
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant)) cttop(Trade & Aid) append
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc    
> if balance==1 & weu==0, robust cluster(country) absorb(country i.year##i.colony_uk i.yea
> r##i.colony_fra i.year##i.colony_ger i.year##i.colony_ussr i.year##i.colony_rest )
(dropped 59 singleton observations)
Warning: time variable yearno has 27 gap(s) in relevant range
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1842
                                                      F(  2,   126) =     7.52
                                                      Prob > F      =   0.0008
Total (centered) SS     =  261095.7929                Centered R2   =  -0.0133
Total (uncentered) SS   =  261095.7929                Uncentered R2 =  -0.0133
Residual SS             =  264567.4108                Root MSE      =    12.34

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5012537   .1902935     2.63   0.009     .1246684     .877839
                |
   lagwe_lngdpc |
            L1. |  -.1897947   .0531135    -3.57   0.001    -.2949048   -.0846846
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.339
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              204.537
                         (Kleibergen-Paap rk Wald F statistic):         36.925
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
------------------------------------------------------------+
        Absorbed FE | Categories  - Redundant  = Num. Coefs |
--------------------+---------------------------------------|
            country |       127         127           0    *|
     year#colony_uk |        56           0          56     |
    year#colony_fra |        54          40          14     |
    year#colony_ger |        52          40          12    ?|
   year#colony_ussr |        45          40           5    ?|
   year#colony_rest |        55          40          15    ?|
------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2  addtext(ColonizerxTime-FE, Yes) addstat(Kleiberge
> n-Paap F-statistic, e( rkf), Countries,e(df_a_nested))  cttop(Colonization) append
coercive_imbalance_tables/Table4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc    
> if balance==1 & weu==0 & colony!=1, robust cluster(country) absorb(country year)
(dropped 16 singleton observations)
Warning: time variable yearno has 26 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =      49                Number of obs =     1098
                                                      F(  2,    48) =     6.46
                                                      Prob > F      =   0.0033
Total (centered) SS     =  204139.4008                Centered R2   =   0.0452
Total (uncentered) SS   =  204139.4008                Uncentered R2 =   0.0452
Residual SS             =  194915.3634                Root MSE      =    13.59

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4559173    .210046     2.17   0.035     .0335916     .878243
                |
   lagwe_lngdpc |
            L1. |  -.1992001   .0598788    -3.33   0.002    -.3195944   -.0788057
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.544
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.959
                         (Kleibergen-Paap rk Wald F statistic):         31.784
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |        49          49           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table4.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Coun
> tries, e(df_a_redundant))  cttop(Colonization) append
coercive_imbalance_tables/Table4.tex
dir : seeout

. 
. 
. //Table 5: Mechanisms 
. *ivreghdfe regime_rest (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc  if 
> balance==1 & weu==0  , robust cluster(country) absorb(country year)
. *outreg2 using "coercive_imbalance_tables/Table5.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2   addstat(Kleibergen-Paap F-statistic, e( rkf), C
> ountries, e(df_a_redundant))    replace NOT INCLUDED IN REPLICATION SINCE BANKS DATA NOT
>  PUBLICLY AVAILABLE
. ivreghdfe v2csreprss (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc  if ba
> lance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1888
                                                      F(  2,   126) =     5.78
                                                      Prob > F      =   0.0040
Total (centered) SS     =  15574985.98                Centered R2   =   0.0496
Total (uncentered) SS   =  15574985.98                Uncentered R2 =   0.0496
Residual SS             =  14803208.21                Root MSE      =    89.55

---------------------------------------------------------------------------------
                |               Robust
     v2csreprss | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   2.637006   1.305756     2.02   0.046     .0529539    5.221059
                |
   lagwe_lngdpc |
            L1. |  -.9887498   .3117246    -3.17   0.002    -1.605644    -.371856
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.908
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.112
                         (Kleibergen-Paap rk Wald F statistic):         39.809
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table5.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2   addstat(Kleibergen-Paap F-statistic, e( rkf), Co
> untries, e(df_a_redundant))    replace
coercive_imbalance_tables/Table5.tex
dir : seeout

. ivreghdfe v2clkill (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc  if bala
> nce==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1888
                                                      F(  2,   126) =     8.49
                                                      Prob > F      =   0.0003
Total (centered) SS     =  13759903.23                Centered R2   =   0.0342
Total (uncentered) SS   =  13759903.23                Uncentered R2 =   0.0342
Residual SS             =  13288630.95                Root MSE      =    84.84

---------------------------------------------------------------------------------
                |               Robust
       v2clkill | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   3.032467   1.467676     2.07   0.041     .1279783    5.936955
                |
   lagwe_lngdpc |
            L1. |  -1.060822   .2693084    -3.94   0.000    -1.593776   -.5278687
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.908
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.112
                         (Kleibergen-Paap rk Wald F statistic):         39.809
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table5.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc) label bdec(2) dec(2) nocons nor2   addstat(Kleibergen-Paap F-statistic, e( rkf), Co
> untries, e(df_a_redundant))   append
coercive_imbalance_tables/Table5.tex
dir : seeout

. ivreghdfe lnmilper (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2)  L.lagwe_lngdpc L.lnmile
> x  if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 18 singleton observations)
Warning: time variable yearno has 16 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     126                Number of obs =     1744
                                                      F(  3,   125) =    25.73
                                                      Prob > F      =   0.0000
Total (centered) SS     =  502.3309409                Centered R2   =  -0.0008
Total (uncentered) SS   =  502.3309409                Uncentered R2 =  -0.0008
Residual SS             =  502.7432078                Root MSE      =    .5437

---------------------------------------------------------------------------------
                |               Robust
       lnmilper | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .0242073   .0103948     2.33   0.021     .0036348    .0447798
                |
   lagwe_lngdpc |
            L1. |   .0019572   .0018395     1.06   0.289    -.0016833    .0055977
                |
        lnmilex |
            L1. |   .3597552   .0453967     7.92   0.000     .2699095    .4496009
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.560
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              179.738
                         (Kleibergen-Paap rk Wald F statistic):         38.492
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lnmilex
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/Table5.tex", keep(L.lagwe_mtech_exf L.lagwe_lng
> dpc L.lnmilex) label bdec(2) dec(2) nocons nor2   addstat(Kleibergen-Paap F-statistic, e
> ( rkf), Countries, e(df_a_redundant) )  append
coercive_imbalance_tables/Table5.tex
dir : seeout

. 
. 
. *###############################################################################
. *       3 ALL REGRESSION TABLES IN THE ONLINE APPENDIX
. *###############################################################################
. preserve 

. keep if balance==1
(4,110 observations deleted)

. outreg2 using "coercive_imbalance_tables/TableA2.tex", label  replace sum(log) keep(lagw
> e_mtech_exf lagwe_mtech_s mtech_exf mtech_sa index_irt_exsa lagwe_mtech_s lagwe_mtech_ex
> f_iv_n2 lagwe_mtech_exf_iv_n2 lagwe_lngdpc  lngdp  v2x_polyarchy democracy v2x_pubcorr v
> 2x_corr v2x_execorr v2clrspct  v2csreprss v2clkill lnmilper n1_democracy_avg n2_democrac
> y_avg n1_interstatewar n2_interstatewar n1_civilwar n2_civilwar distw1_democracy distw1_
> interstatewar distw1_civilwar distw1_lngdp interstatewar_3y civilwar_3y ally_USA ally_Ru
> ssia lntrade lnaid lnmilex) sortvar(lagwe_mtech_exf lagwe_mtech_s mtech_exf mtech_sa ind
> ex_irt_exsa lagwe_mtech_s lagwe_mtech_exf_iv_n2 lagwe_mtech_exf_iv_n2 lagwe_lngdpc  lngd
> p  v2x_polyarchy democracy v2x_pubcorr v2x_corr v2x_execorr v2clrspct  v2csreprss v2clki
> ll lnmilper n1_democracy_avg n2_democracy_avg n1_interstatewar n2_interstatewar n1_civil
> war n2_civilwar distw1_democracy distw1_interstatewar distw1_civilwar distw1_lngdp inter
> statewar_3y civilwar_3y ally_USA ally_Russia lntrade lnaid lnmilex) eqkeep(mean sd N) 

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
        year |      2,378    1966.076    40.03798       1856       2012
       ccode |      2,378    427.9765    244.6779         20        920
  country_id |      2,377    84.23517    56.47343          3        210
   countryno |      2,378    85.74685    50.82688          1        174
         weu |      2,378    .1976451    .3983067          0          1
-------------+---------------------------------------------------------
       mtech |      2,152    20.38476    5.183272          2         29
   mtech_exf |      2,195    11.77677     4.55227          1         20
    mtech_sa |      2,152    8.490706    .9574427          1          9
mtech_exf_~u |      2,195    2.338497    5.084496          0         20
 lagwe_mtech |      2,152    16.13243    25.06521        -24        273
-------------+---------------------------------------------------------
lagwe_mtec~f |      2,195    16.14214     21.9563        -24         94
lagwe_mtec~s |      2,378    8.994533    27.98555        -14        100
lagwe_mtec~u |      2,195    .7908884    4.244521        -20         59
index_ir~xsa |      2,232    .6121817    .6374627  -1.032094   2.287909
index_irt_sa |      2,232    .4451228    .6080907  -2.279462   .8755882
-------------+---------------------------------------------------------
index_irt_~u |      2,232    .1109981    .3888511  -1.032094   2.287909
mtech_exf_~2 |      2,231    11.71701    3.991188   .9691835   17.08065
lagwe_mt~500 |      2,231    16.47548     10.7945          0    44.0957
lagwe_mt~750 |      2,231    16.47589    10.65689          0    44.0957
lagwe_m~1000 |      2,231    16.56263    10.60982          0    44.0957
-------------+---------------------------------------------------------
lagwe_mtec~p |      2,231    15.63022    10.44076          0    45.0315
lagwe_mtec~1 |      2,231    16.46738    11.44004  -.0822058   51.73657
      lngdpc |      2,232    8.168381     1.08383    4.89784   11.56383
lagwe_lngdpc |      2,378    56.15097    45.30349        -70        100
  lagwe_naes |      2,050    161.8337    167.4614        -65        419
-------------+---------------------------------------------------------
lagwe_scho~g |      1,770    62.88475    122.8952        -57        572
   lagwe_ups |      2,367    125.1905    150.7008       -142        612
v2x_polyar~y |      2,371    37.12033    27.49214         .6       92.6
 v2x_pubcorr |      2,364    45.04492    30.35412         .1       97.9
    v2x_corr |      2,362    48.07341    29.86886         .2       96.2
-------------+---------------------------------------------------------
 v2x_execorr |      2,373    46.84534    31.45195         .4       98.1
    v2clkill |      2,377    46.68864    168.8045     -318.1      346.1
   v2clrspct |      2,366    16.43005    155.3603     -353.4      404.9
  v2csreprss |      2,377    25.63664    156.3872     -370.5      330.6
   democracy |      2,377    .3706353    .4830765          0          1
-------------+---------------------------------------------------------
democracy_~s |      2,377    .0046277    .1405693         -1          1
e_total_oi~c |      2,061    252.4809    1657.104          0   40871.99
e_total_re~c |      1,851    390.5651     1832.14          0   41662.14
       milex |      2,246     1705574     5628121          0   1.03e+08
      milper |      2,296    143.8024    373.6047          0       4650
-------------+---------------------------------------------------------
        tpop |      2,290    33364.57    117150.6       1043    1377065
interstate~r |      2,290    .0331878    .1791657          0          1
    civilwar |      2,378    .0702271    .2555831          0          1
civilwar_c~l |      2,290    .0475983    .2129612          0          1
     lntrade |      2,197      7.8089    2.472351  -2.764797    15.3894
-------------+---------------------------------------------------------
    ally_USA |      2,378    .2754415    .4468305          0          1
 ally_Russia |      2,378    .0672834    .2505649          0          1
       lnaid |      2,378    9.652173     9.26398          0   23.00591
      colony |      2,378    .3309504    .4706542          0          1
   colony_we |      2,378    .2682927     .443164          0          1
-------------+---------------------------------------------------------
  colony_ger |      2,378    .0121951    .1097792          0          1
  colony_fra |      2,378     .116905    .3213746          0          1
   colony_uk |      2,378    .1232128    .3287504          0          1
 colony_ussr |      2,378    .0294365    .1690622          0          1
 colony_rest |      2,378    .0391085    .1938939          0          1
-------------+---------------------------------------------------------
         ups |        885    .4966959    .2274426   .0169865          1
         aes |      2,050     .452063    .2603677   .0014925        .95
        naes |      2,050     .547937    .2603677        .05   .9985075
   schooling |      1,796     4.70404    3.197593        .01     12.804
      census |      1,501    .1272485    .3333622          0          1
-------------+---------------------------------------------------------
yearbook_p~d |      1,373    .6504006    .4770172          0          1
   Diplomacy |      2,364    1934.142    46.71014       1817       2005
  University |      2,004    1915.882    1189.573       1088       9999
     lnmilex |      2,243    11.71837    2.662874   4.574711   18.44677
 civilwar_3y |      2,378    .1084945    .3110694          0          1
-------------+---------------------------------------------------------
interstat~3y |      2,293    .0645443    .2457736          0          1
    lnmilper |      2,261     3.71853     1.57031  -1.098612   8.444622
     coldwar |      2,378    .4066442    .4913107          0          1
le_total_o~c |      2,061   -2.647629    5.317136  -6.907755    10.6182
le_total_r~c |      1,851    2.095789    4.530181  -6.907755   10.63735
-------------+---------------------------------------------------------
    r3_iv_n2 |      2,231    .5555046     .357871          0    .928896
    m2_iv_n2 |      2,231    .8905933    .2861315          0          1
       distw |      2,232    975214.1    374869.8   364153.9    2421934
      distw2 |      2,232    970733.4    374433.8   361950.6    2421934
distw2_int~r |      2,232     .043598    .0805898          0   .5303748
-------------+---------------------------------------------------------
distw2_dem~y |      2,232    .3816113    .1581394   .0087075   .7401802
distw2_civ~r |      2,232    .0530785    .0345381          0   .2074214
distw2_ind~a |      2,232    .5009929    .5919949   -1.31142   1.303517
lagwe_mtec~2 |      2,231    16.61623    10.85389          0   45.64799
lagwe_mtec~v |      1,499    20.39808    20.94976  -9.745098         78
-------------+---------------------------------------------------------
lagwe_lngd~v |      1,499    152.7135    178.5602      -48.5        537
n1_interst~r |      2,232    .0985663    .2981456          0          1
n1_democracy |      2,232     .561828    .4962738          0          1
 n1_civilwar |      2,232    .2428315    .4288902          0          1
 n1_ally_USA |      2,232    .3651434    .4815782          0          1
-------------+---------------------------------------------------------
n1_ally_Ru~a |      2,232    .2267025    .4187924          0          1
n1_index_i~a |      2,232    2.357626    2.989069   -13.1593   18.61967
n1_democra~g |      2,232    .3171112    .3644279          0          1
n2_interst~r |      2,232    .1491935     .356359          0          1
n2_democracy |      2,232    .7262545    .4459798          0          1
-------------+---------------------------------------------------------
 n2_civilwar |      2,232    .3723118    .4835292          0          1
n2_index_i~a |      2,232    5.074064    6.642586  -12.17457    34.2342
n2_democra~g |      2,232    .3223144    .3063627          0          1
distw1_int~r |      2,232    .0436433     .084401          0   .6180998
distw1_dem~y |      2,232    .3748026    .1751648   .0220876   .8041152
-------------+---------------------------------------------------------
distw1_civ~r |      2,232     .054951    .0400871          0    .308954
distw1_ind~a |      2,232    .5130923    .5862185  -1.261117   1.372965
distw1_lngdp |      2,232    13.25484     .802973   10.95796   14.71296
adoptiony~s1 |        520    1575.402    118.3235       1428       1888
adoptiony~s2 |        417    1692.962    114.2762       1547       1861
-------------+---------------------------------------------------------
adoptiony~s3 |        523    1734.685    85.85041       1610       1888
adoptiony~s4 |        671    1848.475    13.77903       1830       1882
adoptiony~s5 |        874    1862.365    10.42304       1852       1888
adoptiony~s6 |      1,223    1871.833    5.804365       1858       1890
adoptionye~7 |      1,156    1884.652    10.86032       1865       1904
-------------+---------------------------------------------------------
adoptionye~8 |      1,179    1893.338    4.704509       1886       1904
adoptiony~m1 |        818    1876.006    7.956658       1863       1909
adoptiony~m2 |      1,016    1899.168    8.185908       1889       1931
adoptionye~9 |      1,984    1962.189    6.367209       1946       1984
adoptiony~a1 |        300     1605.02    121.9926       1500       1887
-------------+---------------------------------------------------------
adoptiony~a2 |        853    1865.012    7.427259       1855       1892
adoptiony~a3 |        724    1872.023    8.836192       1855       1896
adoptiony~a4 |        992    1879.622    6.058014       1874       1898
adoptiony~a5 |      1,005    1907.517    4.498666       1901       1926
adoptiony~r1 |      1,773    1934.484    22.57317       1915       2006
-------------+---------------------------------------------------------
adoptiony~r2 |      1,644    1951.505    13.98275       1936       2006
adoptiony~r3 |      1,890    1958.422    11.17432       1946       2008
adoptiony~r4 |      1,794    1964.882    9.806236       1953       2008
adoptiony~r5 |      1,815    1971.074    9.763073       1958       2008
adoptiony~r6 |      1,351    1989.038      8.7324       1973       2009
-------------+---------------------------------------------------------
adoptiony~h1 |      1,086    1990.149    11.36272       1973       2010
adoptiony~h2 |        230    1999.717    6.043738       1990       2007
adoptiony~t1 |      1,813    1944.202     20.7435       1918       1999
adoptiony~t2 |      1,855    1956.137    13.86045       1942       2001
adoptiony~t3 |      1,657    1966.986    14.49595       1950       2007
-------------+---------------------------------------------------------
adoptiony~t4 |      1,288     1978.05    9.590355       1965       2010
adoptiony~t5 |        799    1998.204    8.121551       1981       2010
  firstuses1 |      2,232    1704.127    168.4929       1428       1991
  firstuses2 |      2,232    1801.144    111.6222       1547       1992
  firstuses3 |      2,232    1811.175    97.64476       1610       1992
-------------+---------------------------------------------------------
  firstuses4 |      2,232    1887.633     45.3903       1830       1992
  firstuses5 |      2,232    1894.457    41.95384       1852       1992
  firstuses6 |      2,232    1898.898     38.3334       1858       1992
  firstuses7 |      2,232    1911.384    34.80933       1865       1992
  firstuses8 |      2,232    1917.684    30.87516       1886       1992
-------------+---------------------------------------------------------
  firstusem1 |      2,232    1909.725    38.93044       1863       1992
  firstusem2 |      2,232    1923.177    29.51188       1889       1992
  firstuses9 |      2,232    1963.951    8.506271       1946       1992
  firstusea1 |      2,232    1793.297    158.1118       1500       1992
  firstusea2 |      2,232    1902.958    42.51323       1855       1992
-------------+---------------------------------------------------------
  firstusea3 |      2,232    1908.101    39.42702       1855       1992
  firstusea4 |      2,232    1911.044    36.94395       1874       1992
  firstusea5 |      2,232     1930.63    25.43668       1901       1992
  firstuser1 |      2,028    1936.264    23.33375       1915       2006
  firstuser2 |      2,024    1952.259    13.99224       1936       2006
-------------+---------------------------------------------------------
  firstuser3 |      1,959    1959.259    11.98816       1946       2008
  firstuser4 |      1,863    1965.523    10.37023       1953       2008
  firstuser5 |      1,863    1971.519    10.06272       1958       2008
  firstuser6 |      1,367    1989.061    8.683673       1973       2009
  firstuseh1 |      1,113    1990.135    11.24233       1973       2010
-------------+---------------------------------------------------------
  firstuseh2 |        230    1999.717    6.043738       1990       2007
  firstuset1 |      2,056    1945.839    21.04392       1918       1999
  firstuset2 |      2,030    1956.882     14.4524       1942       2001
  firstuset3 |      1,725    1967.826    14.83655       1950       2007
  firstuset4 |      1,336    1978.519    9.724117       1965       2010
-------------+---------------------------------------------------------
  firstuset5 |        811    1998.097    8.108013       1981       2010
      s1_use |      2,231           1           0          1          1
      s2_use |      2,229    .9968596     .055964          0          1
      s3_use |      2,229    .9968596     .055964          0          1
      s4_use |      2,213    .9950294    .0703431          0          1
-------------+---------------------------------------------------------
      s5_use |      2,212    .9900542    .0992536          0          1
      s6_use |      2,230    .9650224    .1837642          0          1
      s7_use |      2,223    .9388214    .2397115          0          1
      s8_use |      2,231    .9166293    .2765037          0          1
      m1_use |      2,224    .9563849    .2042831          0          1
-------------+---------------------------------------------------------
      m2_use |      2,227    .9052537      .29293          0          1
      s9_use |      2,211    .6313885    .4825374          0          1
      a1_use |      2,231           1           0          1          1
      a2_use |      2,226    .9757412     .153886          0          1
      a3_use |      2,221    .9666817    .1795069          0          1
-------------+---------------------------------------------------------
      a4_use |      2,228    .9510772    .2157551          0          1
      a5_use |      2,226    .8773585     .328099          0          1
      r1_use |      2,231    .7212013    .4485088          0          1
      r2_use |      2,225    .6260674    .4839548          0          1
      r3_use |      2,231    .5459435    .4979963          0          1
-------------+---------------------------------------------------------
      r4_use |      2,231     .474675    .4994702          0          1
      r5_use |      2,231    .4222322    .4940259          0          1
      r6_use |      2,231    .1555356    .3624958          0          1
      h1_use |      2,231    .1532945    .3603519          0          1
      h2_use |      2,231    .0138951     .117082          0          1
-------------+---------------------------------------------------------
      t1_use |      2,231    .6593456    .4740356          0          1
      t2_use |      2,231    .5818019    .4933737          0          1
      t3_use |      2,231      .41013    .4919673          0          1
      t4_use |      2,231    .2223218    .4158995          0          1
      t5_use |      2,231    .0497535    .2174839          0          1
-------------+---------------------------------------------------------
      yearno |      2,378    29.51892     10.0095          2         41
     balance |      2,378           1           0          1          1

Following variable is string, not included:  
iso3  country  region  continent  independent  
coercive_imbalance_tables/TableA2.tex
dir : seeout

. restore

. 
. //Appendix Table A3 -- First Stage Results from 2SLS-Analyses
. reghdfe L.lagwe_mtech_exf L.lagwe_mtech_exf_iv_n2 L.lagwe_lngdpc  if balance==1 & weu==0
>  & v2x_polyarchy!=., vce(cluster country) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 8 iterations)

HDFE Linear regression                            Number of obs   =      1,882
Absorbing 2 HDFE groups                           F(   2,    126) =      20.22
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.8514
                                                  Adj R-squared   =     0.8368
                                                  Within R-sq.    =     0.1277
Number of clusters (country) =        127         Root MSE        =     9.3613

                                       (Std. err. adjusted for 127 clusters in country)
---------------------------------------------------------------------------------------
                      |               Robust
    L.lagwe_mtech_exf | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------------+----------------------------------------------------------------
lagwe_mtech_exf_iv_n2 |
                  L1. |   1.380888     .21876     6.31   0.000     .9479684    1.813808
                      |
         lagwe_lngdpc |
                  L1. |   .0623503   .0333304     1.87   0.064    -.0036095    .1283101
                      |
                _cons |  -9.790549   4.957717    -1.97   0.050    -19.60172    .0206267
---------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA3.tex", keep(L.lagwe_mtech_exf_iv_n2) lab
> el addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Countries, e(df_a_redun
> dant)) cttop(Polyarchy)  replace
coercive_imbalance_tables/TableA3.tex
dir : seeout

. reghdfe L.lagwe_mtech_exf L.lagwe_mtech_exf_iv_n2 L.lagwe_lngdpc  if balance==1 & weu==0
>  & v2x_pubcorr!=., vce(cluster country) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,875
Absorbing 2 HDFE groups                           F(   2,    126) =      20.31
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.8509
                                                  Adj R-squared   =     0.8362
                                                  Within R-sq.    =     0.1289
Number of clusters (country) =        127         Root MSE        =     9.3825

                                       (Std. err. adjusted for 127 clusters in country)
---------------------------------------------------------------------------------------
                      |               Robust
    L.lagwe_mtech_exf | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------------+----------------------------------------------------------------
lagwe_mtech_exf_iv_n2 |
                  L1. |   1.392271   .2200344     6.33   0.000     .9568299    1.827713
                      |
         lagwe_lngdpc |
                  L1. |   .0643754   .0336224     1.91   0.058    -.0021623    .1309132
                      |
                _cons |  -10.14064   5.021385    -2.02   0.046    -20.07781   -.2034672
---------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA3.tex",  keep(L.lagwe_mtech_exf_iv_n2) la
> bel addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Countries, e(df_a_redu
> ndant)) cttop(Public Sector Corruption)  append
coercive_imbalance_tables/TableA3.tex
dir : seeout

. 
. //Appendix Table A4 -- Alternative Instrument Construction
. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 , robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    10.21
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =  -0.0102
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0102
Residual SS             =  286992.0295                Root MSE      =    12.49

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5487931   .1751893     3.13   0.002     .2020988    .8954875
                |
   lagwe_lngdpc |
            L1. |  -.2003625   .0505487    -3.96   0.000     -.300397   -.1003281
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.932
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.378
                         (Kleibergen-Paap rk Wald F statistic):         39.846
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA4.tex", label addtext(Sample, Excl. WEU) 
> bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) cttop(Instrument excl. neighbors)  replace
coercive_imbalance_tables/TableA4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2_500) L.lagwe_lngdpc
>  if balance==1 & weu==0 , robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    10.11
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =  -0.0075
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0075
Residual SS             =  286223.4435                Root MSE      =    12.47

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5422626   .1779728     3.05   0.003     .1900597    .8944656
                |
   lagwe_lngdpc |
            L1. |  -.1997873   .0503234    -3.97   0.000    -.2993757   -.1001988
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.418
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.372
                         (Kleibergen-Paap rk Wald F statistic):         43.227
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2_500
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA4.tex", label addtext(Sample, Excl. WEU) 
> bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) cttop(Instrument excl. countries within 500km)  append
coercive_imbalance_tables/TableA4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2_750) L.lagwe_lngdpc
>  if balance==1 & weu==0 , robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =     8.34
                                                      Prob > F      =   0.0004
Total (centered) SS     =   284096.366                Centered R2   =  -0.0923
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0923
Residual SS             =  310304.9165                Root MSE      =    12.99

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .7087996   .2349912     3.02   0.003     .2437589     1.17384
                |
   lagwe_lngdpc |
            L1. |  -.2144571   .0563589    -3.81   0.000    -.3259896   -.1029245
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             15.113
                                                   Chi-sq(1) P-val =    0.0001
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              135.921
                         (Kleibergen-Paap rk Wald F statistic):         17.171
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2_750
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA4.tex", label addtext(Sample, Excl. WEU) 
> bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) cttop(Instrument excl. countries within 750km)  append
coercive_imbalance_tables/TableA4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2_1000) L.lagwe_lngdp
> c if balance==1 & weu==0 , robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =     7.73
                                                      Prob > F      =   0.0007
Total (centered) SS     =   284096.366                Centered R2   =  -0.1964
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.1964
Residual SS             =  339903.3534                Root MSE      =    13.59

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .8584544   .2804562     3.06   0.003     .3034399    1.413469
                |
   lagwe_lngdpc |
            L1. |  -.2276398   .0615188    -3.70   0.000    -.3493836    -.105896
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             12.901
                                                   Chi-sq(1) P-val =    0.0003
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              116.404
                         (Kleibergen-Paap rk Wald F statistic):         13.486
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2_1000
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA4.tex", label addtext(Sample, Excl. WEU) 
> bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) cttop(Instrument excl. countries within 1000km)  append
coercive_imbalance_tables/TableA4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n1) L.lagwe_lngdpc if 
> balance==1 & weu==0 , robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    10.64
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =   0.0024
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0024
Residual SS             =  283406.3181                Root MSE      =    12.41

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5171942   .1602585     3.23   0.002     .2000473    .8343411
                |
   lagwe_lngdpc |
            L1. |   -.197579   .0498785    -3.96   0.000    -.2962871    -.098871
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             26.167
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              284.544
                         (Kleibergen-Paap rk Wald F statistic):         56.744
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n1
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA4.tex", label addtext(Sample, Excl. WEU) 
> bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) cttop(Instrument includes neighbors)  append
coercive_imbalance_tables/TableA4.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2_pop) L.lagwe_lngdpc
>  if balance==1 & weu==0 , robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    10.03
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =  -0.0007
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0007
Residual SS             =  284303.8438                Root MSE      =    12.43

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5253879   .1734266     3.03   0.003     .1821817    .8685941
                |
   lagwe_lngdpc |
            L1. |  -.1983008     .05015    -3.95   0.000    -.2975462   -.0990554
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             27.005
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              241.472
                         (Kleibergen-Paap rk Wald F statistic):         44.337
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2_pop
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA4.tex", label addtext(Sample, Excl. WEU) 
> bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) cttop(Instrument includes pop. weights)  append
coercive_imbalance_tables/TableA4.tex
dir : seeout

. 
. 
. //Appendix Table A5: Formative Arms Technology Index and Democracy
. reghdfe v2x_polyarchy L.mtech_exf L.mtech_sa L.lngdp  if balance==1, vce(cluster country
> ) absorb(country year)
(dropped 3 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      2,314
Absorbing 2 HDFE groups                           F(   3,    142) =       9.30
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7873
                                                  Adj R-squared   =     0.7688
                                                  Within R-sq.    =     0.0489
Number of clusters (country) =        143         Root MSE        =    13.2784

                              (Std. err. adjusted for 143 clusters in country)
------------------------------------------------------------------------------
             |               Robust
v2x_polyar~y | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
   mtech_exf |
         L1. |   -.933776   .3651126    -2.56   0.012    -1.655535   -.2120174
             |
    mtech_sa |
         L1. |  -.3424971   1.194075    -0.29   0.775    -2.702957    2.017963
             |
      lngdpc |
         L1. |   8.748179   1.849965     4.73   0.000     5.091148    12.40521
             |
       _cons |  -20.14717   18.96708    -1.06   0.290    -57.64151    17.34717
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA5.tex", keep(L.mtech_exf L.mtech_sa L.lng
> dp) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 addstat(Countries, e(df_a_r
> edundant))  sortvar(L.mtech_exf L.mtech_sa L.lngdp) cttop(OLS) replace
coercive_imbalance_tables/TableA5.tex
dir : seeout

. reghdfe v2x_polyarchy L.mtech_exf L.mtech_exf_weu L.lngdp  if balance==1, vce(cluster co
> untry) absorb(country i.weu#i.year year)
(dropped 3 singleton observations)
(MWFE estimator converged in 10 iterations)

HDFE Linear regression                            Number of obs   =      2,368
Absorbing 3 HDFE groups                           F(   3,    142) =      15.47
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.8050
                                                  Adj R-squared   =     0.7846
                                                  Within R-sq.    =     0.0608
Number of clusters (country) =        143         Root MSE        =    12.7614

                               (Std. err. adjusted for 143 clusters in country)
-------------------------------------------------------------------------------
              |               Robust
v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
    mtech_exf |
          L1. |  -1.396884   .3479073    -4.02   0.000    -2.084631   -.7091372
              |
mtech_exf_weu |
          L1. |   2.195763   2.204003     1.00   0.321    -2.161135     6.55266
              |
       lngdpc |
          L1. |   9.172007   1.891601     4.85   0.000      5.43267    12.91134
              |
        _cons |  -26.42089   16.91425    -1.56   0.121    -59.85717    7.015386
-------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
    weu#year |        79           0          79     |
        year |        40          40           0     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA5.tex",  keep(L.mtech_exf L.mtech_exf_weu
>  L.lngdp) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 addstat(Countries,  e
> (df_a_nested))  sortvar(L.mtech_exf L.mtech_sa L.mtech_exf_weu L.lngdp) cttop(OLS) appen
> d 
coercive_imbalance_tables/TableA5.tex
dir : seeout

. reghdfe v2x_polyarchy L.mtech_exf L.lngdp  if balance==1 & weu==0, vce(cluster country) 
> absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,882
Absorbing 2 HDFE groups                           F(   2,    126) =      19.35
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7456
                                                  Adj R-squared   =     0.7206
                                                  Within R-sq.    =     0.0616
Number of clusters (country) =        127         Root MSE        =    12.4749

                              (Std. err. adjusted for 127 clusters in country)
------------------------------------------------------------------------------
             |               Robust
v2x_polyar~y | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
   mtech_exf |
         L1. |   -1.35942   .3442713    -3.95   0.000    -2.040723   -.6781173
             |
      lngdpc |
         L1. |   8.238456   1.999019     4.12   0.000     4.282456    12.19446
             |
       _cons |  -18.61163   16.97235    -1.10   0.275    -52.19941    14.97615
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA5.tex", keep(L.mtech_exf   L.lngdp) label
>    addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Countries, e(df_a_redun
> dant))  sortvar(L.mtech_exf L.mtech_sa L.mtech_exf_weu L.lngdp) cttop(OLS) append  
coercive_imbalance_tables/TableA5.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.mtech_exf =L.mtech_exf_iv_n2) L.lngdp  if balance==1 & weu==
> 0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    11.96
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0101
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0101
Residual SS             =  281239.6239                Root MSE      =    12.36

------------------------------------------------------------------------------
             |               Robust
v2x_polyar~y | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
   mtech_exf |
         L1. |  -3.540946   1.493076    -2.37   0.019      -6.4957   -.5861927
             |
      lngdpc |
         L1. |   9.529478   2.091602     4.56   0.000     5.390258     13.6687
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.840
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              255.206
                         (Kleibergen-Paap rk Wald F statistic):         27.266
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.mtech_exf
Included instruments: L.lngdpc
Excluded instruments: L.mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA5.tex",  keep(L.mtech_exf L.lngdp) label 
> addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-statisti
> c, e( rkf), Countries, e(df_a_redundant)) cttop(2SLS)   append
coercive_imbalance_tables/TableA5.tex
dir : seeout

. xtabond2 v2x_polyarchy L.(mtech_exf lngdpc v2x_polyarchy) i.year if balance==1 & weu==0,
>   gmm(L.(v2x_polyarchy mtech_exf lngdpc), lag(1 2) collapse) iv(i.year) twostep  cluster
> (country)
Favoring speed over space. To switch, type or click on mata: mata set matafavor space, per
> m.
1852b.year dropped due to collinearity
2004.year dropped due to collinearity
Warning: Two-step estimated covariance matrix of moments is singular.
  Using a generalized inverse to calculate optimal weighting matrix for two-step estimatio
> n.
  Difference-in-Sargan/Hansen statistics may be negative.

Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: ccode                           Number of obs      =      1898
Time variable : yearno                          Number of groups   =       153
Number of instruments = 49                      Obs per group: min =         1
Wald chi2(42) = 114417.99                                      avg =     12.41
Prob > chi2   =     0.000                                      max =        40
                                 (Std. err. adjusted for clustering on country)
-------------------------------------------------------------------------------
              |              Corrected
v2x_polyarchy | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
--------------+----------------------------------------------------------------
    mtech_exf |
          L1. |  -.2908535   .3084258    -0.94   0.346     -.895357      .31365
              |
       lngdpc |
          L1. |   2.609349   1.262447     2.07   0.039      .134999      5.0837
              |
v2x_polyarchy |
          L1. |   .8021663    .041538    19.31   0.000     .7207533    .8835793
              |
         year |
        1856  |  -7.082836   4.117995    -1.72   0.085    -15.15396    .9882862
        1860  |  -8.608635   4.434921    -1.94   0.052    -17.30092    .0836511
        1864  |  -7.603564   4.422267    -1.72   0.086    -16.27105     1.06392
        1868  |  -8.345436   4.383434    -1.90   0.057    -16.93681    .2459368
        1872  |    -7.9679   4.513243    -1.77   0.077    -16.81369    .8778938
        1876  |  -6.925907   4.044887    -1.71   0.087    -14.85374    1.001926
        1880  |  -8.122536    3.82325    -2.12   0.034    -15.61597   -.6291032
        1884  |   -7.12979   3.545489    -2.01   0.044    -14.07882   -.1807602
        1888  |  -6.826647   3.451655    -1.98   0.048    -13.59177   -.0615275
        1892  |  -7.069672   3.544065    -1.99   0.046    -14.01591   -.1234326
        1896  |  -6.025092   3.500251    -1.72   0.085    -12.88546    .8352751
        1900  |  -6.873948   3.255963    -2.11   0.035    -13.25552   -.4923774
        1904  |  -7.108496   3.241349    -2.19   0.028    -13.46142   -.7555689
        1908  |  -6.686209   3.198343    -2.09   0.037    -12.95485   -.4175726
        1912  |  -3.941036   2.909469    -1.35   0.176    -9.643491    1.761419
        1916  |  -7.063648   2.753595    -2.57   0.010     -12.4606   -1.666702
        1920  |  -4.446655   2.717905    -1.64   0.102    -9.773651    .8803407
        1924  |  -5.371684   2.551682    -2.11   0.035    -10.37289   -.3704785
        1928  |  -7.298777     2.7394    -2.66   0.008     -12.6679   -1.929652
        1932  |  -7.657404   3.087089    -2.48   0.013    -13.70799   -1.606821
        1936  |  -7.463653   2.737549    -2.73   0.006    -12.82915   -2.098156
        1940  |  -4.345764     3.0919    -1.41   0.160    -10.40578    1.714249
        1944  |  -6.269857   3.084891    -2.03   0.042    -12.31613   -.2235814
        1948  |   3.370688   3.438586     0.98   0.327    -3.368816    10.11019
        1952  |  -5.267073   2.662195    -1.98   0.048    -10.48488   -.0492656
        1956  |  -4.973809   2.531611    -1.96   0.049    -9.935675   -.0119429
        1960  |  -4.039688   2.343935    -1.72   0.085    -8.633716    .5543406
        1964  |  -5.854889   1.813262    -3.23   0.001    -9.408818    -2.30096
        1968  |  -7.322383   2.167421    -3.38   0.001    -11.57045   -3.074316
        1972  |   -7.40939   1.829852    -4.05   0.000    -10.99583   -3.822946
        1976  |  -7.781297   1.790106    -4.35   0.000    -11.28984   -4.272754
        1980  |  -3.892797   1.887039    -2.06   0.039    -7.591325   -.1942697
        1984  |  -4.715316   1.442405    -3.27   0.001    -7.542377   -1.888254
        1988  |  -2.149499   1.788068    -1.20   0.229    -5.654048    1.355051
        1992  |   4.340042   1.991387     2.18   0.029     .4369954    8.243088
        1996  |    2.40975   1.026051     2.35   0.019     .3987261    4.420773
        2000  |  -.0378536   1.030349    -0.04   0.971      -2.0573    1.981593
        2008  |  -.5248577   1.003948    -0.52   0.601     -2.49256    1.442844
        2012  |  -.5784679    .973718    -0.59   0.552     -2.48692    1.329984
              |
        _cons |  -6.933723   8.985904    -0.77   0.440    -24.54577    10.67833
-------------------------------------------------------------------------------
Instruments for first differences equation
  Standard
    D.(1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year)
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(1/2).(L.v2x_polyarchy L.mtech_exf L.lngdpc) collapsed
Instruments for levels equation
  Standard
    1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year
    _cons
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    D.(L.v2x_polyarchy L.mtech_exf L.lngdpc) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -6.19  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =  -0.17  Pr > z =  0.866
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(6)    =   6.11  Prob > chi2 =  0.411
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(6)    =   4.43  Prob > chi2 =  0.619
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
    Hansen test excluding group:     chi2(3)    =   0.90  Prob > chi2 =  0.824
    Difference (null H = exogenous): chi2(3)    =   3.52  Prob > chi2 =  0.318


. outreg2 using "coercive_imbalance_tables/TableA5.tex", keep(L.mtech_exf L.lngdpc L.v2x_p
> olyarchy) label addtext(Sample, Excl. WEU) bdec(2) dec(2)  nocons nor2 noni  addstat(Cou
> ntries, e(N_g) ) cttop(System GMM) append
coercive_imbalance_tables/TableA5.tex
dir : seeout

. 
. 
. 
. //Appendix Table A6: Arms Technology Index (IRT) and Democracy
. reghdfe v2x_polyarchy L.index_irt_exsa L.index_irt_sa L.lngdpc  if balance==1, vce(clust
> er country) absorb(country year)
(dropped 3 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      2,368
Absorbing 2 HDFE groups                           F(   3,    142) =      10.03
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7867
                                                  Adj R-squared   =     0.7686
                                                  Within R-sq.    =     0.0504
Number of clusters (country) =        143         Root MSE        =    13.2271

                                (Std. err. adjusted for 143 clusters in country)
--------------------------------------------------------------------------------
               |               Robust
 v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------+----------------------------------------------------------------
index_irt_exsa |
           L1. |  -7.572066   2.608801    -2.90   0.004    -12.72917    -2.41496
               |
  index_irt_sa |
           L1. |   .0701534   2.019383     0.03   0.972    -3.921786    4.062092
               |
        lngdpc |
           L1. |   8.854742   1.839687     4.81   0.000     5.218029    12.49146
               |
         _cons |  -30.56158   15.01499    -2.04   0.044    -60.24337   -.8797875
--------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA6.tex",  keep(L.index_irt_exsa L.index_ir
> t_sa L.lagwe_mtech_s L.lngdpc) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 
> addstat(Countries, e(df_a_redundant))  sortvar(L.index_irt_exsa L.lagwe_mtech_s L.lngdpc
> ) cttop(OLS) replace
coercive_imbalance_tables/TableA6.tex
dir : seeout

. reghdfe v2x_polyarchy L.index_irt_exsa L.index_irt_exsa_weu L.lngdpc  if balance==1, vce
> (cluster country) absorb(country i.weu#i.year year)
(dropped 3 singleton observations)
(MWFE estimator converged in 10 iterations)

HDFE Linear regression                            Number of obs   =      2,368
Absorbing 3 HDFE groups                           F(   3,    142) =      17.75
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.8059
                                                  Adj R-squared   =     0.7856
                                                  Within R-sq.    =     0.0651
Number of clusters (country) =        143         Root MSE        =    12.7317

                                    (Std. err. adjusted for 143 clusters in country)
------------------------------------------------------------------------------------
                   |               Robust
     v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------------+----------------------------------------------------------------
    index_irt_exsa |
               L1. |  -11.51905   2.624068    -4.39   0.000    -16.70634   -6.331769
                   |
index_irt_exsa_weu |
               L1. |   16.83311   11.17515     1.51   0.134    -5.258057    38.92427
                   |
            lngdpc |
               L1. |   9.538478   1.908969     5.00   0.000     5.764808    13.31215
                   |
             _cons |  -35.49593   15.87762    -2.24   0.027    -66.88298   -4.108872
------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
    weu#year |        79           0          79     |
        year |        40          40           0     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA6.tex",keep(L.index_irt_exsa L.index_irt_
> exsa_weu L.lngdpc) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 addstat(Coun
> tries,  e(df_a_nested))  sortvar(L.index_irt_exsa L.lagwe_mtech_s L.index_irt_exsa_weu L
> .lngdpc) cttop(OLS) append
coercive_imbalance_tables/TableA6.tex
dir : seeout

. reghdfe v2x_polyarchy L.index_irt_exsa L.lngdpc if balance==1 & weu==0, vce(cluster coun
> try) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,882
Absorbing 2 HDFE groups                           F(   2,    126) =      22.52
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7469
                                                  Adj R-squared   =     0.7221
                                                  Within R-sq.    =     0.0666
Number of clusters (country) =        127         Root MSE        =    12.4422

                                (Std. err. adjusted for 127 clusters in country)
--------------------------------------------------------------------------------
               |               Robust
 v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------+----------------------------------------------------------------
index_irt_exsa |
           L1. |  -11.16522   2.605604    -4.29   0.000    -16.32164   -6.008811
               |
        lngdpc |
           L1. |   8.604359    2.01565     4.27   0.000     4.615446    12.59327
               |
         _cons |  -30.51117   16.37642    -1.86   0.065    -62.91962    1.897278
--------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA6.tex", keep(L.index_irt_exsa L.lngdpc) l
> abel   addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Countries,  e(df_a_
> nested))  sortvar(L.index_irt_exsa L.lagwe_mtech_s L.index_irt_exsa_weu L.lngdpc) cttop(
> OLS) append
coercive_imbalance_tables/TableA6.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.index_irt_exsa = L.distw2_index_irt_exsa)  L.lngdpc  if bala
> nce==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  2,   126) =    10.38
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =  -0.0260
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0260
Residual SS             =  291485.6931                Root MSE      =    12.59

--------------------------------------------------------------------------------
               |               Robust
 v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------+----------------------------------------------------------------
index_irt_exsa |
           L1. |  -32.67333   13.47725    -2.42   0.017     -59.3444   -6.002256
               |
        lngdpc |
           L1. |   10.85897   2.434677     4.46   0.000      6.04081    15.67712
--------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             16.788
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              182.279
                         (Kleibergen-Paap rk Wald F statistic):         21.870
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.index_irt_exsa
Included instruments: L.lngdpc
Excluded instruments: L.distw2_index_irt_exsa
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA6.tex", keep(L.index_irt_exsa L.lngdp) la
> bel addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paap F-stat
> istic, e( rkf), Countries, e(df_a_redundant)) cttop(2SLS)   append
coercive_imbalance_tables/TableA6.tex
dir : seeout

. xtabond2 v2x_polyarchy L.v2x_polyarchy L.(index_irt_exsa lngdpc) i.year if balance==1 & 
> weu==0,  gmm(L.(v2x_polyarchy index_irt_exsa lngdpc), lag(1 2) collapse) iv(i.year) twos
> tep  cluster(country)
Favoring speed over space. To switch, type or click on mata: mata set matafavor space, per
> m.
1852b.year dropped due to collinearity
2004.year dropped due to collinearity
Warning: Two-step estimated covariance matrix of moments is singular.
  Using a generalized inverse to calculate optimal weighting matrix for two-step estimatio
> n.
  Difference-in-Sargan/Hansen statistics may be negative.

Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: ccode                           Number of obs      =      1898
Time variable : yearno                          Number of groups   =       153
Number of instruments = 49                      Obs per group: min =         1
Wald chi2(42) = 117069.77                                      avg =     12.41
Prob > chi2   =     0.000                                      max =        40
                                  (Std. err. adjusted for clustering on country)
--------------------------------------------------------------------------------
               |              Corrected
 v2x_polyarchy | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
---------------+----------------------------------------------------------------
 v2x_polyarchy |
           L1. |   .7953743   .0370719    21.45   0.000     .7227148    .8680338
               |
index_irt_exsa |
           L1. |  -5.458901   1.906328    -2.86   0.004    -9.195235   -1.722567
               |
        lngdpc |
           L1. |   2.954035   1.302538     2.27   0.023     .4011085    5.506962
               |
          year |
         1856  |  -14.42896   3.513624    -4.11   0.000    -21.31554   -7.542384
         1860  |  -15.80618   3.780643    -4.18   0.000     -23.2161   -8.396254
         1864  |  -14.23491   3.730287    -3.82   0.000    -21.54614   -6.923682
         1868  |  -14.71402   3.524884    -4.17   0.000    -21.62266    -7.80537
         1872  |  -14.26073   3.657457    -3.90   0.000    -21.42922   -7.092249
         1876  |  -13.56097   3.193778    -4.25   0.000    -19.82066    -7.30128
         1880  |  -13.62476   3.144849    -4.33   0.000    -19.78855   -7.460971
         1884  |  -10.94269    2.95994    -3.70   0.000    -16.74407   -5.141316
         1888  |  -10.86494   2.765238    -3.93   0.000     -16.2847   -5.445169
         1892  |  -11.33638   2.925663    -3.87   0.000    -17.07058   -5.602187
         1896  |  -10.38048   2.839964    -3.66   0.000    -15.94671   -4.814253
         1900  |  -10.73681   2.654973    -4.04   0.000    -15.94046   -5.533155
         1904  |  -11.29974   2.566236    -4.40   0.000    -16.32947   -6.270007
         1908  |  -10.52842   2.541299    -4.14   0.000    -15.50927   -5.547566
         1912  |  -7.542756   2.276725    -3.31   0.001    -12.00505   -3.080458
         1916  |  -10.06067   2.167111    -4.64   0.000    -14.30813   -5.813209
         1920  |  -7.240056   2.106585    -3.44   0.001    -11.36889   -3.111225
         1924  |  -7.639311   1.938176    -3.94   0.000    -11.43806   -3.840556
         1928  |  -9.409056   2.229121    -4.22   0.000    -13.77805   -5.040058
         1932  |  -9.510089   2.272337    -4.19   0.000    -13.96379   -5.056391
         1936  |  -10.25617    2.22546    -4.61   0.000    -14.61799   -5.894352
         1940  |  -5.519484   2.392709    -2.31   0.021    -10.20911   -.8298595
         1944  |   -7.84224   2.374924    -3.30   0.001    -12.49701   -3.187474
         1948  |   1.282996    2.77441     0.46   0.644    -4.154747    6.720739
         1952  |  -5.772121   2.167072    -2.66   0.008     -10.0195   -1.524737
         1956  |  -4.999136    2.09592    -2.39   0.017    -9.107063   -.8912083
         1960  |  -5.177906   1.865664    -2.78   0.006     -8.83454   -1.521272
         1964  |  -6.962681     1.4538    -4.79   0.000    -9.812075   -4.113286
         1968  |  -8.987101   1.767841    -5.08   0.000      -12.452   -5.522196
         1972  |  -8.936828   1.549783    -5.77   0.000    -11.97435   -5.899309
         1976  |  -9.323026   1.543595    -6.04   0.000    -12.34842   -6.297636
         1980  |    -4.9265   1.612366    -3.06   0.002    -8.086678   -1.766321
         1984  |  -5.854656   1.357281    -4.31   0.000    -8.514878   -3.194434
         1988  |  -2.410158   1.626956    -1.48   0.139    -5.598934    .7786173
         1992  |   4.045552   1.925783     2.10   0.036     .2710868    7.820017
         1996  |   2.067828   1.024602     2.02   0.044     .0596446    4.076011
         2000  |  -.0426233   1.055786    -0.04   0.968    -2.111926     2.02668
         2008  |   -.290579   1.026463    -0.28   0.777     -2.30241    1.721252
         2012  |  -.1524875   .9887793    -0.15   0.877    -2.090459    1.785484
               |
         _cons |  -8.226782   9.734269    -0.85   0.398     -27.3056    10.85203
--------------------------------------------------------------------------------
Instruments for first differences equation
  Standard
    D.(1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year)
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(1/2).(L.v2x_polyarchy L.index_irt_exsa L.lngdpc) collapsed
Instruments for levels equation
  Standard
    1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year
    _cons
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    D.(L.v2x_polyarchy L.index_irt_exsa L.lngdpc) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -6.32  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =  -0.17  Pr > z =  0.865
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(6)    =   4.68  Prob > chi2 =  0.586
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(6)    =   6.81  Prob > chi2 =  0.339
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
    Hansen test excluding group:     chi2(3)    =   0.91  Prob > chi2 =  0.822
    Difference (null H = exogenous): chi2(3)    =   5.89  Prob > chi2 =  0.117


. outreg2 using "coercive_imbalance_tables/TableA6.tex",  keep(L.index_irt_exsa  L.lngdpc)
>  label addtext(Sample, Excl. WEU) bdec(2) dec(2)  nocons nor2 noni  addstat(Countries, e
> (N_g) ) cttop(System GMM) append
coercive_imbalance_tables/TableA6.tex
dir : seeout

. 
. //Appendix Table A7: Arms Technology and Democracy -- Using the BMR Indicator
. replace democracy=democracy*100
(1,032 real changes made)

. reghdfe democracy L.lagwe_mtech_exf L.lagwe_mtech_s L.lagwe_lngdpc  if balance==1, vce(c
> luster country) absorb(country year)
(dropped 3 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      2,374
Absorbing 2 HDFE groups                           F(   3,    142) =      12.88
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.6369
                                                  Adj R-squared   =     0.6062
                                                  Within R-sq.    =     0.0439
Number of clusters (country) =        143         Root MSE        =    30.3154

                                 (Std. err. adjusted for 143 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
      democracy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4605228   .1122467     4.10   0.000     .2386323    .6824133
                |
  lagwe_mtech_s |
            L1. |   .0670127    .084407     0.79   0.429     -.099844    .2338693
                |
   lagwe_lngdpc |
            L1. |  -.2826621   .0616379    -4.59   0.000    -.4045085   -.1608157
                |
          _cons |   45.56405   3.936438    11.57   0.000     37.78246    53.34564
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA7.tex", keep(L.lagwe_mtech_exf L.lagwe_mt
> ech_s L.lagwe_lngdpc) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 addstat(C
> ountries, e(df_a_redundant))  sortvar(L.lagwe_mtech_exf L.lagwe_mtech_s L.lagwe_lngdpc) 
> cttop(OLS) replace
coercive_imbalance_tables/TableA7.tex
dir : seeout

. reghdfe democracy L.lagwe_mtech_exf L.lagwe_mtech_exf_weu L.lagwe_lngdpc  if balance==1,
>  vce(cluster country) absorb(country i.weu#i.year year)
(dropped 3 singleton observations)
(MWFE estimator converged in 10 iterations)

HDFE Linear regression                            Number of obs   =      2,374
Absorbing 3 HDFE groups                           F(   3,    142) =      13.57
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.6518
                                                  Adj R-squared   =     0.6155
                                                  Within R-sq.    =     0.0427
Number of clusters (country) =        143         Root MSE        =    29.9562

                                     (Std. err. adjusted for 143 clusters in country)
-------------------------------------------------------------------------------------
                    |               Robust
          democracy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------------+----------------------------------------------------------------
    lagwe_mtech_exf |
                L1. |    .485141   .1110472     4.37   0.000     .2656217    .7046603
                    |
lagwe_mtech_exf_weu |
                L1. |  -.2705831   .5164137    -0.52   0.601    -1.291435    .7502692
                    |
       lagwe_lngdpc |
                L1. |  -.2887992   .0615244    -4.69   0.000    -.4104213   -.1671772
                    |
              _cons |   45.92898   3.931255    11.68   0.000     38.15764    53.70033
-------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       143         143           0    *|
    weu#year |        79           0          79     |
        year |        40          40           0     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA7.tex", keep(L.lagwe_mtech_exf L.lagwe_mt
> ech_exf_weu L.lagwe_lngdpc) label   addtext(Sample, Full) bdec(2) dec(2) nocons nor2 add
> stat(Countries,  e(df_a_nested))  sortvar(L.lagwe_mtech_exf L.lagwe_mtech_s L.lagwe_mtec
> h_exf_weu L.lagwe_lngdpc) cttop(OLS) append
coercive_imbalance_tables/TableA7.tex
dir : seeout

. reghdfe democracy L.lagwe_mtech_exf L.lagwe_lngdpc  if balance==1 & weu==0, vce(cluster 
> country) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,888
Absorbing 2 HDFE groups                           F(   2,    126) =      18.45
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.5883
                                                  Adj R-squared   =     0.5480
                                                  Within R-sq.    =     0.0476
Number of clusters (country) =        127         Root MSE        =    30.1127

                                 (Std. err. adjusted for 127 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
      democracy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4920749   .1099953     4.47   0.000     .2743975    .7097523
                |
   lagwe_lngdpc |
            L1. |  -.3152984   .0738027    -4.27   0.000    -.4613517   -.1692451
                |
          _cons |     40.555   5.602569     7.24   0.000     29.46768    51.64232
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA7.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label   addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Countries,  
> e(df_a_nested))  sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) cttop(OLS) append
coercive_imbalance_tables/TableA7.tex
dir : seeout

. ivreghdfe democracy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc  if bal
> ance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1888
                                                      F(  2,   126) =    12.72
                                                      Prob > F      =   0.0000
Total (centered) SS     =  1636651.786                Centered R2   =   0.0211
Total (uncentered) SS   =  1636651.786                Uncentered R2 =   0.0211
Residual SS             =  1602168.141                Root MSE      =    29.46

---------------------------------------------------------------------------------
                |               Robust
      democracy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   1.000293   .3703553     2.70   0.008      .267371    1.733216
                |
   lagwe_lngdpc |
            L1. |  -.3606296   .0753808    -4.78   0.000     -.509806   -.2114532
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.840
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              220.168
                         (Kleibergen-Paap rk Wald F statistic):         39.666
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA7.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label addtext(Sample, Excl. WEU) bdec(2) dec(2) nocons nor2 addstat(Kleibergen-Paa
> p F-statistic, e( rkf), Countries, e(df_a_redundant)) cttop(2SLS)   append
coercive_imbalance_tables/TableA7.tex
dir : seeout

. xtabond2 democracy  L.(democracy lagwe_mtech_exf lagwe_lngdpc) if balance==1 & weu==0,  
> gmm(L.(democracy lagwe_mtech_exf lagwe_lngdpc), lag(1 2) collapse) iv(i.year) twostep ro
> bust  cluster(country)
Favoring speed over space. To switch, type or click on mata: mata set matafavor space, per
> m.
Warning: Two-step estimated covariance matrix of moments is singular.
  Using a generalized inverse to calculate optimal weighting matrix for two-step estimatio
> n.
  Difference-in-Sargan/Hansen statistics may be negative.

Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: ccode                           Number of obs      =      1907
Time variable : yearno                          Number of groups   =       153
Number of instruments = 49                      Obs per group: min =         1
Wald chi2(3)  =   1687.24                                      avg =     12.46
Prob > chi2   =     0.000                                      max =        40
                                   (Std. err. adjusted for clustering on country)
---------------------------------------------------------------------------------
                |              Corrected
      democracy | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
----------------+----------------------------------------------------------------
      democracy |
            L1. |   .7502585   .0423955    17.70   0.000     .6671648    .8333521
                |
lagwe_mtech_exf |
            L1. |   .1953021   .0583638     3.35   0.001     .0809112     .309693
                |
   lagwe_lngdpc |
            L1. |  -.2129802   .0712472    -2.99   0.003    -.3526221   -.0733383
                |
          _cons |   19.20684   5.916918     3.25   0.001     7.609895    30.80379
---------------------------------------------------------------------------------
Instruments for first differences equation
  Standard
    D.(1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year)
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(1/2).(L.democracy L.lagwe_mtech_exf L.lagwe_lngdpc) collapsed
Instruments for levels equation
  Standard
    1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year
    1880.year 1884.year 1888.year 1892.year 1896.year 1900.year 1904.year
    1908.year 1912.year 1916.year 1920.year 1924.year 1928.year 1932.year
    1936.year 1940.year 1944.year 1948.year 1952.year 1956.year 1960.year
    1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.year
    1992.year 1996.year 2000.year 2004.year 2008.year 2012.year
    _cons
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    D.(L.democracy L.lagwe_mtech_exf L.lagwe_lngdpc) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -5.17  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =   1.34  Pr > z =  0.181
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(45)   =  87.42  Prob > chi2 =  0.000
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(45)   =  64.24  Prob > chi2 =  0.031
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
    Hansen test excluding group:     chi2(42)   =  61.96  Prob > chi2 =  0.024
    Difference (null H = exogenous): chi2(3)    =   2.28  Prob > chi2 =  0.515
  gmm(L.democracy L.lagwe_mtech_exf L.lagwe_lngdpc, collapse lag(1 2))
    Hansen test excluding group:     chi2(36)   =  48.28  Prob > chi2 =  0.083
    Difference (null H = exogenous): chi2(9)    =  15.96  Prob > chi2 =  0.068
  iv(1852b.year 1856.year 1860.year 1864.year 1868.year 1872.year 1876.year 1880.year 1884
> .year 1888.year 1892.year 1896.year 1900.year 1904.year 1908.year 1912.year 1916.year 19
> 20.year 1924.year 1928.year 1932.year 1936.year 1940.year 1944.year 1948.year 1952.year 
> 1956.year 1960.year 1964.year 1968.year 1972.year 1976.year 1980.year 1984.year 1988.yea
> r 1992.year 1996.year 2000.year 2004.year 2008.year 2012.year)
    Hansen test excluding group:     chi2(6)    =   7.65  Prob > chi2 =  0.265
    Difference (null H = exogenous): chi2(39)   =  56.60  Prob > chi2 =  0.034


. outreg2 using "coercive_imbalance_tables/TableA7.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label addtext(Sample, Excl. WEU) bdec(2) dec(2)  nocons nor2 noni  addstat(Countri
> es, e(N_g) ) cttop(System GMM) append
coercive_imbalance_tables/TableA7.tex
dir : seeout

. 
. //Appendix Table A8: Robustness of 2SLS-Results for Democracy (Identical to Table 4 abov
> e, which does not show controls)
. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.n
> 1_democracy_avg L.n2_democracy_avg if balance==1 & weu==0, robust cluster(country) absor
> b(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  4,   126) =    12.38
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0366
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0366
Residual SS             =  273687.3575                Root MSE      =     12.2

----------------------------------------------------------------------------------
                 |               Robust
   v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |   .4668069   .1853742     2.52   0.013     .0999568    .8336569
                 |
    lagwe_lngdpc |
             L1. |  -.1987217   .0481884    -4.12   0.000     -.294085   -.1033584
                 |
n1_democracy_avg |
             L1. |   7.646955   3.623861     2.11   0.037       .47544    14.81847
                 |
n2_democracy_avg |
             L1. |   2.736598    5.34315     0.51   0.609    -7.837338    13.31053
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.398
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              191.250
                         (Kleibergen-Paap rk Wald F statistic):         37.092
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.n1_democracy_avg L.n2_democracy_avg
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e( 
> rkf), Countries, e(df_a_redundant)) replace  
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.n
> 1_interstatewar L.n2_interstatewar L.n1_civilwar L.n2_civilwar  if balance==1 & weu==0, 
> robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  6,   126) =     5.82
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0007
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0007
Residual SS             =  283904.1803                Root MSE      =    12.44

----------------------------------------------------------------------------------
                 |               Robust
   v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |   .5260894   .1758369     2.99   0.003     .1781132    .8740655
                 |
    lagwe_lngdpc |
             L1. |  -.2003176    .050039    -4.00   0.000    -.2993432   -.1012919
                 |
n1_interstatewar |
             L1. |   .7088044   1.255897     0.56   0.573    -1.776579    3.194188
                 |
n2_interstatewar |
             L1. |    1.22631   .9935615     1.23   0.219    -.7399188    3.192539
                 |
     n1_civilwar |
             L1. |   -.836317   .9408749    -0.89   0.376    -2.698281    1.025647
                 |
     n2_civilwar |
             L1. |  -.0694333   .9813033    -0.07   0.944    -2.011404    1.872537
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.432
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              210.293
                         (Kleibergen-Paap rk Wald F statistic):         37.503
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.n1_interstatewar L.n2_interstatewar
                      L.n1_civilwar L.n2_civilwar
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex",  label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(d
> f_a_redundant)) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc L.n1_democracy_avg L.n2_democra
> cy_avg L.n1_interstatewar L.n2_interstatewar L.n1_civilwar L.n2_civilwar)  append 
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.d
> istw1_democracy L.distw1_interstatewar L.distw1_civilwar L.distw1_lngdp  if balance==1 &
>  weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  6,   126) =     5.10
                                                      Prob > F      =   0.0001
Total (centered) SS     =   284096.366                Centered R2   =   0.0161
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0161
Residual SS             =  279528.7114                Root MSE      =    12.34

--------------------------------------------------------------------------------------
                     |               Robust
       v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
     lagwe_mtech_exf |
                 L1. |    .540632   .2023748     2.67   0.009     .1401382    .9411259
                     |
        lagwe_lngdpc |
                 L1. |  -.1717192   .0459544    -3.74   0.000    -.2626615   -.0807768
                     |
    distw1_democracy |
                 L1. |   4.421407    18.3755     0.24   0.810    -31.94317    40.78598
                     |
distw1_interstatewar |
                 L1. |   .3506999   8.533305     0.04   0.967    -16.53646    17.23786
                     |
     distw1_civilwar |
                 L1. |  -1.181058    14.7407    -0.08   0.936    -30.35246    27.99035
                     |
        distw1_lngdp |
                 L1. |  -24.28214   10.47803    -2.32   0.022    -45.01785   -3.546433
--------------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             20.339
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              151.070
                         (Kleibergen-Paap rk Wald F statistic):         28.752
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.distw1_democracy L.distw1_interstatewar
                      L.distw1_civilwar L.distw1_lngdp
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex",  label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(d
> f_a_redundant)) append  
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.i
> nterstatewar_3y L.civilwar_3y  if balance==1 & weu==0, robust cluster(country) absorb(co
> untry year) 
(dropped 18 singleton observations)
Warning: time variable yearno has 20 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1812
                                                      F(  4,   126) =     6.42
                                                      Prob > F      =   0.0001
Total (centered) SS     =  275213.9494                Centered R2   =  -0.0278
Total (uncentered) SS   =  275213.9494                Uncentered R2 =  -0.0278
Residual SS             =  282860.1368                Root MSE      =    12.65

----------------------------------------------------------------------------------
                 |               Robust
   v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |   .6270155   .1876416     3.34   0.001     .2556782    .9983527
                 |
    lagwe_lngdpc |
             L1. |  -.2067142   .0505232    -4.09   0.000    -.3066981   -.1067304
                 |
interstatewar_3y |
             L1. |  -1.511121   1.447874    -1.04   0.299     -4.37642    1.354178
                 |
     civilwar_3y |
             L1. |  -.3211159   1.307116    -0.25   0.806    -2.907859    2.265628
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.509
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              197.264
                         (Kleibergen-Paap rk Wald F statistic):         39.632
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.interstatewar_3y L.civilwar_3y
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Sample, Excl. WEU) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant))  append
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.a
> lly_USA L.ally_Russia if balance==1 & weu==0, robust cluster(country) absorb(country yea
> r)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  4,   126) =    12.11
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =   0.0578
Total (uncentered) SS   =   284096.366                Uncentered R2 =   0.0578
Residual SS             =  267669.9545                Root MSE      =    12.07

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |    .481112    .159482     3.02   0.003     .1655017    .7967223
                |
   lagwe_lngdpc |
            L1. |  -.1957571   .0434456    -4.51   0.000    -.2817348   -.1097795
                |
       ally_USA |
            L1. |   12.89828   3.319328     3.89   0.000     6.329426    19.46713
                |
    ally_Russia |
            L1. |  -.7243786   2.092215    -0.35   0.730     -4.86481    3.416053
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.097
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              218.836
                         (Kleibergen-Paap rk Wald F statistic):         37.199
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.ally_USA L.ally_Russia
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex",   label bdec(2) dec(2) nocons nor
> 2 addtext(Sample, Excl. WEU) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(
> df_a_redundant))  append
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> ntrade    if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 18 singleton observations)
Warning: time variable yearno has 11 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     123                Number of obs =     1726
                                                      F(  3,   122) =     7.01
                                                      Prob > F      =   0.0002
Total (centered) SS     =  262723.6001                Centered R2   =  -0.0531
Total (uncentered) SS   =  262723.6001                Uncentered R2 =  -0.0531
Residual SS             =  276667.6807                Root MSE      =     12.8

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |    .701445   .2043462     3.43   0.001     .2969212    1.105969
                |
   lagwe_lngdpc |
            L1. |  -.1495033   .0468768    -3.19   0.002    -.2423006    -.056706
                |
        lntrade |
            L1. |   2.638049   1.141815     2.31   0.023     .3777121    4.898385
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.500
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              188.070
                         (Kleibergen-Paap rk Wald F statistic):         37.035
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lntrade
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       123         123           0    *|
        year |        35           0          35     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex",  label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(d
> f_a_redundant)) append
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> naid    if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1882
                                                      F(  3,   126) =    17.13
                                                      Prob > F      =   0.0000
Total (centered) SS     =   284096.366                Centered R2   =  -0.0020
Total (uncentered) SS   =   284096.366                Uncentered R2 =  -0.0020
Residual SS             =  284652.0618                Root MSE      =    12.44

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5394168   .1779868     3.03   0.003     .1871861    .8916475
                |
   lagwe_lngdpc |
            L1. |  -.2100204   .0454857    -4.62   0.000    -.3000352   -.1200056
                |
          lnaid |
            L1. |   .2159095   .2176896     0.99   0.323    -.2148918    .6467108
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.068
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              219.402
                         (Kleibergen-Paap rk Wald F statistic):         39.804
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lnaid
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Sample, Excl. WEU) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df
> _a_redundant)) append
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc    
> if balance==1 & weu==0, robust cluster(country) absorb(country i.year##i.colony_uk i.yea
> r##i.colony_fra i.year##i.colony_ger i.year##i.colony_ussr i.year##i.colony_rest )
(dropped 59 singleton observations)
Warning: time variable yearno has 27 gap(s) in relevant range
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1842
                                                      F(  2,   126) =     7.52
                                                      Prob > F      =   0.0008
Total (centered) SS     =  261095.7929                Centered R2   =  -0.0133
Total (uncentered) SS   =  261095.7929                Uncentered R2 =  -0.0133
Residual SS             =  264567.4108                Root MSE      =    12.34

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .5012537   .1902935     2.63   0.009     .1246684     .877839
                |
   lagwe_lngdpc |
            L1. |  -.1897947   .0531135    -3.57   0.001    -.2949048   -.0846846
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.339
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              204.537
                         (Kleibergen-Paap rk Wald F statistic):         36.925
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
------------------------------------------------------------+
        Absorbed FE | Categories  - Redundant  = Num. Coefs |
--------------------+---------------------------------------|
            country |       127         127           0    *|
     year#colony_uk |        56           0          56     |
    year#colony_fra |        54          40          14     |
    year#colony_ger |        52          40          12    ?|
   year#colony_ussr |        45          40           5    ?|
   year#colony_rest |        55          40          15    ?|
------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2 addtext(Sample, Excl. WEU, ColonizerxTime-FE, Yes
> ) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a_nested))  append
coercive_imbalance_tables/TableA8.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc    
> if balance==1 & weu==0 & colony!=1, robust cluster(country) absorb(country year)
(dropped 16 singleton observations)
Warning: time variable yearno has 26 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =      49                Number of obs =     1098
                                                      F(  2,    48) =     6.46
                                                      Prob > F      =   0.0033
Total (centered) SS     =  204139.4008                Centered R2   =   0.0452
Total (uncentered) SS   =  204139.4008                Uncentered R2 =   0.0452
Residual SS             =  194915.3634                Root MSE      =    13.59

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4559173    .210046     2.17   0.035     .0335916     .878243
                |
   lagwe_lngdpc |
            L1. |  -.1992001   .0598788    -3.33   0.002    -.3195944   -.0788057
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.544
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.959
                         (Kleibergen-Paap rk Wald F statistic):         31.784
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |        49          49           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA8.tex", keep(L.lagwe_mtech_exf L.lagwe_ln
> gdpc) label bdec(2) dec(2) nocons nor2 addtext(Sample, Excl. WEU & ex-colonies) addstat(
> Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a_redundant)) append
coercive_imbalance_tables/TableA8.tex
dir : seeout

. 
. *Table A.9: Excluding Regions and Time Periods
. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 & region!="Sub-Saharan Africa", robust cluster(country) absorb(count
> ry year)
(dropped 59 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =      87                Number of obs =     1437
                                                      F(  2,    86) =     8.87
                                                      Prob > F      =   0.0003
Total (centered) SS     =  246853.6897                Centered R2   =   0.0419
Total (uncentered) SS   =  246853.6897                Uncentered R2 =   0.0419
Residual SS             =  236521.8939                Root MSE      =    13.02

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4946545    .179647     2.75   0.007     .1375279     .851781
                |
   lagwe_lngdpc |
            L1. |  -.2107013    .054433    -3.87   0.000    -.3189106   -.1024921
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             22.387
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              240.435
                         (Kleibergen-Paap rk Wald F statistic):         36.891
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |        87          87           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, 2SLS) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))cttop(Excl. SSA) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) replace
coercive_imbalance_tables/TableA9.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 & colony!=1, robust cluster(country) absorb(country year)
(dropped 16 singleton observations)
Warning: time variable yearno has 26 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =      49                Number of obs =     1098
                                                      F(  2,    48) =     6.46
                                                      Prob > F      =   0.0033
Total (centered) SS     =  204139.4008                Centered R2   =   0.0452
Total (uncentered) SS   =  204139.4008                Uncentered R2 =   0.0452
Residual SS             =  194915.3634                Root MSE      =    13.59

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4559173    .210046     2.17   0.035     .0335916     .878243
                |
   lagwe_lngdpc |
            L1. |  -.1992001   .0598788    -3.33   0.002    -.3195944   -.0788057
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.544
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.959
                         (Kleibergen-Paap rk Wald F statistic):         31.784
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |        49          49           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, 2SLS) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))cttop(Excl. excolonies) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA9.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 & region!="MENA", robust cluster(country) absorb(country year)
(dropped 40 singleton observations)
Warning: time variable yearno has 25 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     106                Number of obs =     1581
                                                      F(  2,   105) =     6.64
                                                      Prob > F      =   0.0019
Total (centered) SS     =  252901.9188                Centered R2   =   0.0130
Total (uncentered) SS   =  252901.9188                Uncentered R2 =   0.0130
Residual SS             =   249614.069                Root MSE      =    12.74

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |    .463937   .2261407     2.05   0.043     .0155417    .9123323
                |
   lagwe_lngdpc |
            L1. |  -.2265246   .0643482    -3.52   0.001    -.3541152    -.098934
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.457
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              147.930
                         (Kleibergen-Paap rk Wald F statistic):         28.256
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       106         106           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, 2SLS) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))cttop(Excl. MENA) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA9.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 & region!="Asia and Oceania", robust cluster(country) absorb(country
>  year)
(dropped 45 singleton observations)
Warning: time variable yearno has 22 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     101                Number of obs =     1476
                                                      F(  2,   100) =     5.97
                                                      Prob > F      =   0.0036
Total (centered) SS     =  206661.9612                Centered R2   =   0.0017
Total (uncentered) SS   =  206661.9612                Uncentered R2 =   0.0017
Residual SS             =  206313.3072                Root MSE      =    11.99

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4899417   .1849435     2.65   0.009      .123019    .8568644
                |
   lagwe_lngdpc |
            L1. |  -.1273446   .0460865    -2.76   0.007     -.218779   -.0359103
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.150
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              190.976
                         (Kleibergen-Paap rk Wald F statistic):         37.126
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       101         101           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, 2SLS) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant)) cttop(Excl. Asia & Oceania) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) appen
> d
coercive_imbalance_tables/TableA9.tex
dir : seeout

. reghdfe v2x_polyarchy  L.c.lagwe_mtech_exf##c.coldwar   L.c.lagwe_lngdpc##c.coldwar if b
> alance==1 & weu==0, vce (cluster country) absorb(country year)
(dropped 19 singleton observations)
note: coldwar is probably collinear with the fixed effects (all partialled-out values are 
> close to zero; tol = 1.0e-09)
note: coldwar is probably collinear with the fixed effects (all partialled-out values are 
> close to zero; tol = 1.0e-09)
(MWFE estimator converged in 8 iterations)
note: coldwar omitted because of collinearity
note: coldwar omitted because of collinearity

HDFE Linear regression                            Number of obs   =      1,882
Absorbing 2 HDFE groups                           F(   4,    126) =       9.48
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7473
                                                  Adj R-squared   =     0.7222
                                                  Within R-sq.    =     0.0681
Number of clusters (country) =        127         Root MSE        =    12.4389

                                     (Std. err. adjusted for 127 clusters in country)
-------------------------------------------------------------------------------------
                    |               Robust
      v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------------+----------------------------------------------------------------
    lagwe_mtech_exf |
                L1. |   .2096753   .0601967     3.48   0.001     .0905477    .3288028
                    |
            coldwar |          0  (omitted)
                    |
 cL.lagwe_mtech_exf#|
          c.coldwar |  -.0358709   .0668238    -0.54   0.592    -.1681133    .0963714
                    |
       lagwe_lngdpc |
                L1. |  -.1924422   .0473253    -4.07   0.000    -.2860975   -.0987869
                    |
            coldwar |          0  (omitted)
                    |
    cL.lagwe_lngdpc#|
          c.coldwar |   .0525925   .0427443     1.23   0.221    -.0319973    .1371822
                    |
              _cons |   39.19914   3.627782    10.81   0.000     32.01987    46.37842
-------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, OLS) addstat(Countries,  e(df_a_redundant)) cttop(Cold War Interactio
> n) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA9.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 & year<1948  | balance==1 & weu==0 &  year>1990 , robust cluster(cou
> ntry) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 48 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1082
                                                      F(  2,   126) =     2.71
                                                      Prob > F      =   0.0702
Total (centered) SS     =  94855.62674                Centered R2   =   0.0160
Total (uncentered) SS   =  94855.62674                Uncentered R2 =   0.0160
Residual SS             =  93335.77029                Root MSE      =    9.424

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .2872029   .3016911     0.95   0.343    -.3098349    .8842408
                |
   lagwe_lngdpc |
            L1. |    -.17019    .073056    -2.33   0.021    -.3147656   -.0256144
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             13.926
                                                   Chi-sq(1) P-val =    0.0002
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              122.743
                         (Kleibergen-Paap rk Wald F statistic):         24.073
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        29           0          29     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, 2SLS) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))cttop(Excl. 1948-91) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA9.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc if 
> balance==1 & weu==0 & year<1960  | balance==1 & weu==0 &  year>1989 , robust cluster(cou
> ntry) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 76 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1186
                                                      F(  2,   126) =     6.96
                                                      Prob > F      =   0.0014
Total (centered) SS     =  131561.7469                Centered R2   =   0.0236
Total (uncentered) SS   =  131561.7469                Uncentered R2 =   0.0236
Residual SS             =  128454.2521                Root MSE      =    10.56

---------------------------------------------------------------------------------
                |               Robust
  v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .4415884    .214683     2.06   0.042     .0167371    .8664398
                |
   lagwe_lngdpc |
            L1. |  -.1889483   .0561863    -3.36   0.001    -.3001393   -.0777573
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.527
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              206.586
                         (Kleibergen-Paap rk Wald F statistic):         35.213
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        32           0          32     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA9.tex", label bdec(2) dec(2) nocons nor2 
> addtext(Estimator, 2SLS) addstat(Kleibergen-Paap F-statistic, e( rkf), Countries, e(df_a
> _redundant))cttop(Excl. 1960-91) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA9.tex
dir : seeout

. 
. *Table A10: Controlling for OIL and Natural Resources
. reghdfe v2x_polyarchy  L.lagwe_mtech_exf  L.lagwe_lngdpc L.le_total_oil_income_pc if bal
> ance==1 & weu==0, vce(cluster country) absorb(country year)
(dropped 1 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,721
Absorbing 2 HDFE groups                           F(   3,    125) =      11.52
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7392
                                                  Adj R-squared   =     0.7111
                                                  Within R-sq.    =     0.0748
Number of clusters (country) =        126         Root MSE        =    12.3955

                                        (Std. err. adjusted for 126 clusters in country)
----------------------------------------------------------------------------------------
                       |               Robust
         v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
       lagwe_mtech_exf |
                   L1. |   .1954472   .0521163     3.75   0.000     .0923025    .2985919
                       |
          lagwe_lngdpc |
                   L1. |  -.1860427   .0509364    -3.65   0.000    -.2868522   -.0852333
                       |
le_total_oil_income_pc |
                   L1. |  -.3966729   .3022944    -1.31   0.192    -.9949511    .2016053
                       |
                 _cons |   38.59576   3.659559    10.55   0.000     31.35304    45.83848
----------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        39           0          39     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA10.tex", label bdec(2) dec(2) nocons nor2
>  addstat(Countries,  e(df_a_redundant)) cttop(OLS) sortvar(L.lagwe_mtech_exf L.lagwe_lng
> dpc) replace
coercive_imbalance_tables/TableA10.tex
dir : seeout

. reghdfe v2x_polyarchy  L.lagwe_mtech_exf  L.lagwe_lngdpc L.le_total_resources_income_pc 
> if balance==1 & weu==0, vce(cluster country) absorb(country year)
(dropped 1 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,587
Absorbing 2 HDFE groups                           F(   3,    125) =      12.46
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7360
                                                  Adj R-squared   =     0.7074
                                                  Within R-sq.    =     0.0636
Number of clusters (country) =        126         Root MSE        =    12.6002

                                         (Std. err. adjusted for 126 clusters in country)
-----------------------------------------------------------------------------------------
                        |               Robust
          v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
------------------------+----------------------------------------------------------------
        lagwe_mtech_exf |
                    L1. |   .2231668   .0483464     4.62   0.000     .1274832    .3188503
                        |
           lagwe_lngdpc |
                    L1. |   -.163068   .0428853    -3.80   0.000    -.2479433   -.0781926
                        |
le_total_resources_in~c |
                    L1. |  -.1652212   .2333132    -0.71   0.480     -.626977    .2965345
                        |
                  _cons |   38.50254   3.299053    11.67   0.000     31.97331    45.03178
-----------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        27           0          27     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA10.tex", label bdec(2) dec(2) nocons nor2
>  addstat(Countries,  e(df_a_redundant)) cttop(OLS) sortvar(L.lagwe_mtech_exf L.lagwe_lng
> dpc) append
coercive_imbalance_tables/TableA10.tex
dir : seeout

. reghdfe v2x_polyarchy  L.lagwe_mtech_exf  L.lagwe_lngdpc L.le_total_oil_income_pc L.le_t
> otal_resources_income_pc if balance==1 & weu==0, vce(cluster country) absorb(country yea
> r)
(dropped 1 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,587
Absorbing 2 HDFE groups                           F(   4,    125) =       9.57
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7373
                                                  Adj R-squared   =     0.7086
                                                  Within R-sq.    =     0.0682
Number of clusters (country) =        126         Root MSE        =    12.5737

                                         (Std. err. adjusted for 126 clusters in country)
-----------------------------------------------------------------------------------------
                        |               Robust
          v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
------------------------+----------------------------------------------------------------
        lagwe_mtech_exf |
                    L1. |   .2145823   .0497927     4.31   0.000     .1160363    .3131283
                        |
           lagwe_lngdpc |
                    L1. |  -.1707402   .0450659    -3.79   0.000    -.2599312   -.0815492
                        |
 le_total_oil_income_pc |
                    L1. |  -.4502602   .3647278    -1.23   0.219    -1.172102    .2715814
                        |
le_total_resources_in~c |
                    L1. |   .0563172   .2626688     0.21   0.831    -.4635371    .5761714
                        |
                  _cons |   37.91292   3.109611    12.19   0.000     31.75861    44.06723
-----------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        27           0          27     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA10.tex", label bdec(2) dec(2) nocons nor2
>  addstat(Countries,  e(df_a_redundant)) cttop(OLS) sortvar(L.lagwe_mtech_exf L.lagwe_lng
> dpc) append
coercive_imbalance_tables/TableA10.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> e_total_oil_income_pc if balance==1 & weu==0 , robust cluster(country) absorb(country ye
> ar)
(dropped 1 singleton observations)
Warning: time variable yearno has 38 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     126                Number of obs =     1721
                                                      F(  3,   125) =     6.80
                                                      Prob > F      =   0.0003
Total (centered) SS     =  257916.3528                Centered R2   =  -0.0112
Total (uncentered) SS   =  257916.3528                Uncentered R2 =  -0.0112
Residual SS             =  260801.8833                Root MSE      =    12.46

----------------------------------------------------------------------------------------
                       |               Robust
         v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------------+----------------------------------------------------------------
       lagwe_mtech_exf |
                   L1. |   .5918678   .1961511     3.02   0.003     .2036605    .9800751
                       |
          lagwe_lngdpc |
                   L1. |  -.2124711   .0556009    -3.82   0.000    -.3225121   -.1024301
                       |
le_total_oil_income_pc |
                   L1. |  -.1549699   .3542115    -0.44   0.662    -.8559984    .5460585
----------------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.573
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              174.103
                         (Kleibergen-Paap rk Wald F statistic):         39.524
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.le_total_oil_income_pc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        39           0          39     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA10.tex", label bdec(2) dec(2) nocons nor2
>  addstat(Kleibergen-Paap F-statistic, e( rkf),Countries,  e(df_a_redundant)) cttop(2SLS)
>  sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA10.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> e_total_resources_income_pc if balance==1 & weu==0 , robust cluster(country) absorb(coun
> try year) 
(dropped 1 singleton observations)
Warning: time variable yearno has 21 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     126                Number of obs =     1587
                                                      F(  3,   125) =     6.73
                                                      Prob > F      =   0.0003
Total (centered) SS     =  242626.2604                Centered R2   =  -0.0406
Total (uncentered) SS   =  242626.2604                Uncentered R2 =  -0.0406
Residual SS             =  252474.9982                Root MSE      =    12.73

-----------------------------------------------------------------------------------------
                        |               Robust
          v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
------------------------+----------------------------------------------------------------
        lagwe_mtech_exf |
                    L1. |   .6555602   .2033227     3.22   0.002     .2531594    1.057961
                        |
           lagwe_lngdpc |
                    L1. |  -.1917261   .0477073    -4.02   0.000    -.2861448   -.0973073
                        |
le_total_resources_in~c |
                    L1. |   .0036979   .2581554     0.01   0.989    -.5072236    .5146194
-----------------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.815
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              163.949
                         (Kleibergen-Paap rk Wald F statistic):         40.391
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.le_total_resources_income_pc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        27           0          27     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA10.tex", label bdec(2) dec(2) nocons nor2
>  addstat(Kleibergen-Paap F-statistic, e( rkf),Countries,  e(df_a_redundant)) cttop(2SLS)
>  sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA10.tex
dir : seeout

. ivreghdfe v2x_polyarchy  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.l
> e_total_oil_income_pc L.le_total_resources_income_pc if balance==1 & weu==0 , robust clu
> ster(country) absorb(country year)
(dropped 1 singleton observations)
Warning: time variable yearno has 21 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     126                Number of obs =     1587
                                                      F(  4,   125) =     5.25
                                                      Prob > F      =   0.0006
Total (centered) SS     =  242626.2604                Centered R2   =  -0.0360
Total (uncentered) SS   =  242626.2604                Uncentered R2 =  -0.0360
Residual SS             =   251348.911                Root MSE      =    12.71

-----------------------------------------------------------------------------------------
                        |               Robust
          v2x_polyarchy | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
------------------------+----------------------------------------------------------------
        lagwe_mtech_exf |
                    L1. |   .6487863   .2067658     3.14   0.002      .239571    1.058002
                        |
           lagwe_lngdpc |
                    L1. |  -.1958153   .0491664    -3.98   0.000    -.2931217   -.0985089
                        |
 le_total_oil_income_pc |
                    L1. |  -.2479452   .3877057    -0.64   0.524    -1.015263    .5193725
                        |
le_total_resources_in~c |
                    L1. |   .1248931   .2655996     0.47   0.639    -.4007614    .6505476
-----------------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.593
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              161.159
                         (Kleibergen-Paap rk Wald F statistic):         40.694
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.le_total_oil_income_pc
                      L.le_total_resources_income_pc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        27           0          27     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA10.tex", label bdec(2) dec(2) nocons nor2
>  addstat(Kleibergen-Paap F-statistic, e( rkf),Countries,  e(df_a_redundant)) cttop(2SLS)
>  sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc) append
coercive_imbalance_tables/TableA10.tex
dir : seeout

. 
. // Table A11: Robustness of 2SLS-results for Bureaucracy 
. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.n1_
> democracy_avg L.n2_democracy_avg if balance==1 & weu==0, robust cluster(country) absorb(
> country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  4,   126) =     1.36
                                                      Prob > F      =   0.2503
Total (centered) SS     =  313810.0553                Centered R2   =  -0.1087
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.1087
Residual SS             =  347921.9324                Root MSE      =    13.78

----------------------------------------------------------------------------------
                 |               Robust
     v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |  -.5894435   .2814549    -2.09   0.038    -1.146434   -.0324525
                 |
    lagwe_lngdpc |
             L1. |   .0826003    .056669     1.46   0.147     -.029546    .1947466
                 |
n1_democracy_avg |
             L1. |   1.000922   3.351841     0.30   0.766    -5.632273    7.634116
                 |
n2_democracy_avg |
             L1. |   4.095056     6.8421     0.60   0.551     -9.44526    17.63537
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.540
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              192.157
                         (Kleibergen-Paap rk Wald F statistic):         37.255
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.n1_democracy_avg L.n2_democracy_avg
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE,    ) addstat(Kleibergen-Paap F-statistic,
>  e( rkf), Countries, e(df_a_redundant))  replace
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.n1_
> interstatewar L.n2_interstatewar L.n1_civilwar L.n2_civilwar  if balance==1 & weu==0, ro
> bust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  6,   126) =     1.72
                                                      Prob > F      =   0.1222
Total (centered) SS     =  313810.0553                Centered R2   =  -0.0951
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.0951
Residual SS             =  343648.7588                Root MSE      =    13.71

----------------------------------------------------------------------------------
                 |               Robust
     v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |  -.5658153    .273832    -2.07   0.041    -1.107721   -.0239098
                 |
    lagwe_lngdpc |
             L1. |   .0831042   .0567617     1.46   0.146    -.0292255    .1954339
                 |
n1_interstatewar |
             L1. |  -1.810091    1.31233    -1.38   0.170    -4.407154    .7869725
                 |
n2_interstatewar |
             L1. |   .8392731   1.275813     0.66   0.512    -1.685523    3.364069
                 |
     n1_civilwar |
             L1. |  -1.764476   1.182893    -1.49   0.138    -4.105386    .5764337
                 |
     n2_civilwar |
             L1. |  -.7154244   1.146808    -0.62   0.534    -2.984924    1.554075
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.531
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              212.091
                         (Kleibergen-Paap rk Wald F statistic):         37.711
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.n1_interstatewar L.n2_interstatewar
                      L.n1_civilwar L.n2_civilwar
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e(
>  rkf), Countries, e(df_a_redundant)) sortvar(L.lagwe_mtech_exf L.lagwe_lngdpc L.n1_democ
> racy_avg L.n2_democracy_avg L.n1_interstatewar L.n2_interstatewar L.n1_civilwar L.n2_civ
> ilwar)   append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.dis
> tw1_democracy L.distw1_interstatewar L.distw1_civilwar L.distw1_lngdp  if balance==1 & w
> eu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  6,   126) =     1.08
                                                      Prob > F      =   0.3786
Total (centered) SS     =  313810.0553                Centered R2   =  -0.1308
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.1308
Residual SS             =  354866.1884                Root MSE      =    13.93

--------------------------------------------------------------------------------------
                     |               Robust
         v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
     lagwe_mtech_exf |
                 L1. |  -.6318295   .3336028    -1.89   0.061     -1.29202    .0283606
                     |
        lagwe_lngdpc |
                 L1. |   .0842393   .0492823     1.71   0.090     -.013289    .1817676
                     |
    distw1_democracy |
                 L1. |    9.69229    20.1955     0.48   0.632    -30.27401    49.65859
                     |
distw1_interstatewar |
                 L1. |   8.002202   10.62453     0.75   0.453    -13.02343    29.02783
                     |
     distw1_civilwar |
                 L1. |   -8.97738   16.95866    -0.53   0.597    -42.53807    24.58331
                     |
        distw1_lngdp |
                 L1. |   1.693219   20.57074     0.08   0.935    -39.01566     42.4021
--------------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             20.320
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              151.542
                         (Kleibergen-Paap rk Wald F statistic):         28.672
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.distw1_democracy L.distw1_interstatewar
                      L.distw1_civilwar L.distw1_lngdp
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e(
>  rkf), Countries, e(df_a_redundant)) append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.int
> erstatewar_3y L.civilwar_3y  if balance==1 & weu==0, robust cluster(country) absorb(coun
> try year)
(dropped 18 singleton observations)
Warning: time variable yearno has 21 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1815
                                                      F(  4,   126) =     2.25
                                                      Prob > F      =   0.0678
Total (centered) SS     =  295046.3933                Centered R2   =  -0.1530
Total (uncentered) SS   =  295046.3933                Uncentered R2 =  -0.1530
Residual SS             =  340191.1295                Root MSE      =    13.86

----------------------------------------------------------------------------------
                 |               Robust
     v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
 lagwe_mtech_exf |
             L1. |  -.7255964   .2801538    -2.59   0.011    -1.280013   -.1711803
                 |
    lagwe_lngdpc |
             L1. |   .0727721    .054332     1.34   0.183    -.0347494    .1802935
                 |
interstatewar_3y |
             L1. |  -.6751489   1.618844    -0.42   0.677    -3.878793    2.528495
                 |
     civilwar_3y |
             L1. |   1.000883   1.294541     0.77   0.441    -1.560975     3.56274
----------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.555
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              198.432
                         (Kleibergen-Paap rk Wald F statistic):         39.846
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.interstatewar_3y L.civilwar_3y
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e(
>  rkf), Countries, e(df_a_redundant))  append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.all
> y_USA L.ally_Russia if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  4,   126) =     4.43
                                                      Prob > F      =   0.0022
Total (centered) SS     =  313810.0553                Centered R2   =  -0.0772
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.0772
Residual SS             =   338041.497                Root MSE      =    13.59

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.5721322   .2540274    -2.25   0.026    -1.074845   -.0694195
                |
   lagwe_lngdpc |
            L1. |    .085159   .0553646     1.54   0.127     -.024406     .194724
                |
       ally_USA |
            L1. |  -1.662803   4.905829    -0.34   0.735    -11.37129    8.045688
                |
    ally_Russia |
            L1. |  -12.23664   4.431219    -2.76   0.007    -21.00589   -3.467387
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.195
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              220.100
                         (Kleibergen-Paap rk Wald F statistic):         37.417
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.ally_USA L.ally_Russia
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e(
>  rkf), Countries, e(df_a_redundant))  append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.lnt
> rade    if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 18 singleton observations)
Warning: time variable yearno has 11 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     123                Number of obs =     1730
                                                      F(  3,   122) =     2.90
                                                      Prob > F      =   0.0379
Total (centered) SS     =  285806.2554                Centered R2   =  -0.2101
Total (uncentered) SS   =  285806.2554                Uncentered R2 =  -0.2101
Residual SS             =  345867.3091                Root MSE      =     14.3

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.8638576   .2983629    -2.90   0.004    -1.454497   -.2732184
                |
   lagwe_lngdpc |
            L1. |   .0042591   .0490169     0.09   0.931    -.0927748     .101293
                |
        lntrade |
            L1. |  -3.826981   1.795827    -2.13   0.035    -7.381999   -.2719629
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             24.498
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              187.821
                         (Kleibergen-Paap rk Wald F statistic):         36.997
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lntrade
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       123         123           0    *|
        year |        35           0          35     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e(
>  rkf), Countries, e(df_a_redundant))  append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc L.lna
> id    if balance==1 & weu==0, robust cluster(country) absorb(country year)
(dropped 19 singleton observations)
Warning: time variable yearno has 29 gap(s) in relevant range
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1875
                                                      F(  3,   126) =     1.68
                                                      Prob > F      =   0.1738
Total (centered) SS     =  313810.0553                Centered R2   =  -0.0986
Total (uncentered) SS   =  313810.0553                Uncentered R2 =  -0.0986
Residual SS             =  344746.0058                Root MSE      =    13.72

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.5656542   .2661748    -2.13   0.036    -1.092406   -.0389022
                |
   lagwe_lngdpc |
            L1. |    .079857   .0554854     1.44   0.153    -.0299469    .1896609
                |
          lnaid |
            L1. |   .0926095   .2279639     0.41   0.685    -.3585245    .5437434
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             25.147
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              221.062
                         (Kleibergen-Paap rk Wald F statistic):         40.003
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc L.lnaid
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex",label bdec(2) dec(2) nocons nor2 
> addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e( 
> rkf), Countries, e(df_a_redundant)) append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc    if
>  balance==1 & weu==0, robust cluster(country) absorb(country i.year##i.colony_uk i.year#
> #i.colony_fra i.year##i.colony_ger i.year##i.colony_ussr i.year##i.colony_rest )
(dropped 59 singleton observations)
Warning: time variable yearno has 28 gap(s) in relevant range
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =     127                Number of obs =     1835
                                                      F(  2,   126) =     2.05
                                                      Prob > F      =   0.1329
Total (centered) SS     =  284511.8721                Centered R2   =  -0.1430
Total (uncentered) SS   =  284511.8721                Uncentered R2 =  -0.1430
Residual SS             =  325202.6392                Root MSE      =    13.71

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.5616673   .2913544    -1.93   0.056    -1.138249    .0149145
                |
   lagwe_lngdpc |
            L1. |   .0805387   .0578218     1.39   0.166    -.0338889    .1949663
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             23.429
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              206.169
                         (Kleibergen-Paap rk Wald F statistic):         37.093
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
------------------------------------------------------------+
        Absorbed FE | Categories  - Redundant  = Num. Coefs |
--------------------+---------------------------------------|
            country |       127         127           0    *|
     year#colony_uk |        56           0          56     |
    year#colony_fra |        54          40          14     |
    year#colony_ger |        52          40          12    ?|
   year#colony_ussr |        45          40           5    ?|
   year#colony_rest |        55          40          15    ?|
------------------------------------------------------------+
? = number of redundant parameters may be higher
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F-statistic, e(
>  rkf), Countries, e(df_a_redundant))  append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. ivreghdfe v2x_pubcorr  (L.lagwe_mtech_exf =L.lagwe_mtech_exf_iv_n2) L.lagwe_lngdpc    if
>  balance==1 & weu==0 & colony!=1, robust cluster(country) absorb(country year)
(dropped 16 singleton observations)
Warning: time variable yearno has 27 gap(s) in relevant range
(MWFE estimator converged in 7 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on country

Number of clusters (country) =      49                Number of obs =     1091
                                                      F(  2,    48) =     1.70
                                                      Prob > F      =   0.1930
Total (centered) SS     =  166193.2506                Centered R2   =  -0.1490
Total (uncentered) SS   =  166193.2506                Uncentered R2 =  -0.1490
Residual SS             =  190952.9869                Root MSE      =    13.49

---------------------------------------------------------------------------------
                |               Robust
    v2x_pubcorr | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |  -.6105859   .3342814    -1.83   0.074    -1.282704    .0615319
                |
   lagwe_lngdpc |
            L1. |   .0640369   .0601528     1.06   0.292    -.0569085    .1849823
---------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             19.581
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):              222.699
                         (Kleibergen-Paap rk Wald F statistic):         31.744
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         L.lagwe_mtech_exf
Included instruments: L.lagwe_lngdpc
Excluded instruments: L.lagwe_mtech_exf_iv_n2
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |        49          49           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA11.tex", label bdec(2) dec(2) nocons nor2
>  addtext(Sample, Excl. WEU & ex-colonies, ColonizerxTime-FE, ) addstat(Kleibergen-Paap F
> -statistic, e( rkf), Countries, e(df_a_redundant))  append
coercive_imbalance_tables/TableA11.tex
dir : seeout

. 
. // Table A12: Mechanisms OLS 
. *reghdfe regime_rest L.lagwe_mtech_exf L.lagwe_lngdpc  if balance==1 & weu==0  , vce(clu
> ster country) absorb(country year)
. *outreg2 using "coercive_imbalance_tables/TableA12.tex", label bdec(2) dec(2) nocons add
> stat(Countries,  e(df_a_redundant)) addtext(Sample, Excl. WEU) replace NOT INCLUDED IN R
> EPLICATION SINCE BANKS DATA NOT PUBLICLY AVAILABLE
. reghdfe v2csreprss L.lagwe_mtech_exf   L.lagwe_lngdpc  if balance==1 & weu==0, vce(clust
> er country) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,888
Absorbing 2 HDFE groups                           F(   2,    126) =      13.12
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.6364
                                                  Adj R-squared   =     0.6009
                                                  Within R-sq.    =     0.0560
Number of clusters (country) =        127         Root MSE        =    92.4822

                                 (Std. err. adjusted for 127 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
     v2csreprss | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   1.864035   .4490211     4.15   0.000      .975436    2.752635
                |
   lagwe_lngdpc |
            L1. |  -.9199804   .3022474    -3.04   0.003    -1.518119   -.3218417
                |
          _cons |   22.82587   23.15507     0.99   0.326    -22.99734    68.64907
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA12.tex", label bdec(2) dec(2) nocons adds
> tat(Countries,  e(df_a_redundant)) addtext(Sample, Excl. WEU) replace
coercive_imbalance_tables/TableA12.tex
dir : seeout

. reghdfe v2clkill L.lagwe_mtech_exf  L.lagwe_lngdpc  if balance==1 & weu==0, vce(cluster 
> country) absorb(country year)
(dropped 19 singleton observations)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =      1,888
Absorbing 2 HDFE groups                           F(   2,    126) =      14.05
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7092
                                                  Adj R-squared   =     0.6807
                                                  Within R-sq.    =     0.0570
Number of clusters (country) =        127         Root MSE        =    86.8796

                                 (Std. err. adjusted for 127 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
       v2clkill | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   1.668437   .4422119     3.77   0.000     .7933129    2.543562
                |
   lagwe_lngdpc |
            L1. |  -.9394677   .2630077    -3.57   0.001    -1.459952   -.4189832
                |
          _cons |   38.94132   20.93049     1.86   0.065    -2.479508    80.36216
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       127         127           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA12.tex", bdec(2) dec(2) nocons addstat(Co
> untries,  e(df_a_redundant)) addtext(Sample, Excl. WEU) append
coercive_imbalance_tables/TableA12.tex
dir : seeout

. reghdfe lnmilper L.lagwe_mtech_exf L.lagwe_lngdpc L.lnmilex  if balance==1 & weu==0, vce
> (cluster country) absorb(country year)
(dropped 18 singleton observations)
(MWFE estimator converged in 8 iterations)

HDFE Linear regression                            Number of obs   =      1,744
Absorbing 2 HDFE groups                           F(   3,    125) =      26.81
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.9084
                                                  Adj R-squared   =     0.8986
                                                  Within R-sq.    =     0.1509
Number of clusters (country) =        126         Root MSE        =     0.5204

                                 (Std. err. adjusted for 126 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
       lnmilper | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
lagwe_mtech_exf |
            L1. |   .0005872   .0025914     0.23   0.821    -.0045416     .005716
                |
   lagwe_lngdpc |
            L1. |   .0027603   .0017358     1.59   0.114    -.0006751    .0061956
                |
        lnmilex |
            L1. |   .3009714   .0350536     8.59   0.000     .2315959    .3703469
                |
          _cons |    .023655   .4772456     0.05   0.961    -.9208732    .9681833
---------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     country |       126         126           0    *|
        year |        40           0          40     |
-----------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation

. outreg2 using "coercive_imbalance_tables/TableA12.tex", bdec(2) dec(2) nocons addstat(Co
> untries,  e(df_a_redundant)) addtext(Sample, Excl. WEU) append
coercive_imbalance_tables/TableA12.tex
dir : seeout

. 
. *###############################################################################
. *       4 ALL FIGURES IN THE MAIN TEXT
. *###############################################################################
. 
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. // FIGURE 1: Effectiveness of Small Arms Technologies. 
. *Source: Zabecki, D.T., 2014. Germany at War: 400 Years of Military History [4 volumes]:
>  400 Years of Military History. ABC-CLIO.
. preserve

. clear

. set obs 9
Number of observations (_N) was 0, now 9.

. 
. gen techname=""
(9 missing values generated)

. gen effective_range=.
(9 missing values generated)

. gen rate_of_fire=.
(9 missing values generated)

. 
. replace techname="Matchlock musket" if _n==1
variable techname was str1 now str16
(1 real change made)

. replace techname="Snaphaunce" if _n==2
(1 real change made)

. replace techname="Flintlock" if _n==3
(1 real change made)

. replace techname="Percussion lock" if _n==4
(1 real change made)

. replace techname="Minié bullet rifle" if _n==5
variable techname was str16 now str19
(1 real change made)

. replace techname="Breechloading rifle" if _n==6
(1 real change made)

. replace techname="Tubular magazine rifle" if _n==7
variable techname was str19 now str22
(1 real change made)

. replace techname="Box magazine rifle" if _n==8
(1 real change made)

. replace techname="Assault rifle" if _n==9
(1 real change made)

. 
. ** Data from Zabecki 2014 (see main text). 
. ** A few corrections for reloading times and expected misfire is made to make the number
> s comparable, see the comments below
. 
. *a. Effective range (in meters)
. replace effective_range=50 if techname=="Matchlock musket"
(1 real change made)

. replace effective_range=50 if techname=="Snaphaunce"
(1 real change made)

. replace effective_range=75 if techname=="Flintlock"
(1 real change made)

. replace effective_range=75 if techname=="Percussion lock"
(1 real change made)

. replace effective_range=300 if techname=="Minié bullet rifle"
(1 real change made)

. replace effective_range=450 if techname=="Breechloading rifle"
(1 real change made)

. replace effective_range=500 if techname=="Tubular magazine rifle"
(1 real change made)

. replace effective_range=500 if techname=="Box magazine rifle"
(1 real change made)

. replace effective_range=400 if techname=="Assault rifle"
(1 real change made)

. 
. *b. Rate of fire (shots/minute)
. replace rate_of_fire=0.5 if techname=="Matchlock musket"
(1 real change made)

. replace rate_of_fire=2 if techname=="Snaphaunce"
(1 real change made)

. replace rate_of_fire=3 if techname=="Flintlock" //I have subtracted 1 shot/minute to tak
> e misfire into account.
(1 real change made)

. replace rate_of_fire=4 if techname=="Percussion lock"
(1 real change made)

. replace rate_of_fire=4 if techname=="Minié bullet rifle"
(1 real change made)

. replace rate_of_fire=10 if techname=="Breechloading rifle"
(1 real change made)

. replace rate_of_fire=15 if techname=="Tubular magazine rifle"
(1 real change made)

. replace rate_of_fire=20 if techname=="Box magazine rifle" //changed to 20 from 17.5 -> 2
> 0 round trench magazines used in WWI, Before that I assume that 10 round magazine could 
> be fired and changed once
(1 real change made)

. replace rate_of_fire=90 if techname=="Assault rifle" //changed to 90 -> practical rate o
> f fire assuming magazine replaced twice in a minute. The rate of fire was otherwise list
> ed as 550 shots/min
(1 real change made)

. 
. *geometric mean
. gen simpleeffect=sqrt(effective_range*rate_of_fire)

. 
. graph hbar simpleeffect, over(techname, sort(simpleeffect)) ///
> ytitle("Geometric mean of range and rate of fire") plotregion(fcolor(white)) graphregion
> (color(white)) ///
> bar(1, color(gs8))

. 
. graph export "coercive_imbalance_figures/Figure1.eps", replace
file coercive_imbalance_figures/Figure1.eps saved as EPS format

. graph export "coercive_imbalance_figures/Figure1.pdf", replace
file coercive_imbalance_figures/Figure1.pdf saved as PDF format

. 
. restore

. 
. // FIGURE 2: How Much Are Other Countries Behind Western Europe in Arms Technology and F
> our Indicators of Modernization 
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. preserve

. drop if year<1850
(5,190 observations deleted)

. drop if year>2010
(346 observations deleted)

. 
. keep if region!="Western Europe and off-shoots" 
(3,703 observations deleted)

. keep if democracy!=.
(13,230 observations deleted)

. 
. collapse (mean) lagwe_*, by(year)

. 
. sort year

. twoway line lagwe_mtech year , lcolor(black) lpattern(solid) || ///
> line lagwe_lngdpc year , lcolor(gs6) lpattern(solid) || ///
> line lagwe_naes year if lagwe_naes<=400, lcolor(gs6) lpattern(longdash)  || ///
> line lagwe_ups year , lcolor(gs6) lpattern(dash_dot) || ///
> line lagwe_schooling year ,lcolor(gs6) lpattern(dot)  ///
> legend(label(1 "Arms technology") label(2 "GDP per capita") label(3 "Industrialization")
>   label(4 "Urbanization rate") label(5 "Years of schooling")) ///
> ytitle("Lag to Western Europe (years)") xtitle("Year") plotregion(fcolor(white)) graphre
> gion(color(white)) scheme(s1color) yline(0)

. 
. graph export "coercive_imbalance_figures/Figure2.eps", replace
file coercive_imbalance_figures/Figure2.eps saved as EPS format

. graph export "coercive_imbalance_figures/Figure2.pdf", replace
file coercive_imbalance_figures/Figure2.pdf saved as PDF format

. 
. restore

. 
. // FIGURE 3: Lags in Arms Technology and Economic Development Across Six World Regions
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. *WEU
. twoway (scatter lagwe_mtech_av lagwe_lngdpc_av if year==2010 & region=="Western Europe a
> nd off-shoots", mlabel(iso3) mlabcolor(black) mlabposition(12) msymbol(i) mcolor(black) 
> legend(off) xtitle(Lag GDP per capita) ytitle(Lag arms technology) title(WEU and offshoo
> ts) ) (lfit lagwe_mtech lagwe_mtech  if year==2010 & region=="Western Europe and off-sho
> ots", lpattern(dash) lcolor(black) range(-50 100) msymbol(iso3) msymbol(i) legend(off) )
>  , scheme(s1color) yscale(range(-50 0 50 100)) ylabel(-50 0 50 100) name(f1)

. *EEU
. twoway (scatter lagwe_mtech_av lagwe_lngdpc_av if year==2010 & region=="Eastern Europe a
> nd Balkan", mlabel(iso3) mlabcolor(black) mlabposition(12) msymbol(i) mcolor(black) lege
> nd(off) xtitle(Lag GDP per capita) ytitle(Lag arms technology) title(Eastern Europe) ) (
> lfit lagwe_mtech lagwe_mtech  if year==2010 & region=="Eastern Europe and Balkan", lpatt
> ern(dash) lcolor(black) range(0 100) msymbol(iso3) msymbol(i) legend(off) ) , scheme(s1c
> olor) yscale(range(0 20 40 60 80 100)) ylabel(0 20 40 60 80 100) name(f2)

. *LAC
. twoway (scatter lagwe_mtech_av lagwe_lngdpc_av if year==2010 & region=="Latin America", 
> mlabel(iso3) mlabcolor(black) mlabposition(12) msymbol(i) mcolor(black) legend(off) xtit
> le(Lag GDP per capita) ytitle(Lag arms technology) title(Latin America) ) (lfit lagwe_mt
> ech lagwe_mtech  if year==2010 & region=="Latin America", lpattern(dash) lcolor(black) r
> ange(0 100) msymbol(iso3) msymbol(i) legend(off) ) , scheme(s1color) yscale(range(0 20 4
> 0 60 80 100)) ylabel(0 20 40 60 80 100) name(f3)

. *ASIA
. twoway (scatter lagwe_mtech_av lagwe_lngdpc_av if year==2010 & region=="Asia and Oceania
> ", mlabel(iso3) mlabcolor(black) mlabposition(12) msymbol(i) mcolor(black) legend(off) x
> title(Lag GDP per capita) ytitle(Lag arms technology) title(Asia) ) (lfit lagwe_mtech la
> gwe_mtech  if year==2010 & region=="Asia and Oceania", lpattern(dash) lcolor(black) rang
> e(0 100) msymbol(iso3) msymbol(i) legend(off) ) , scheme(s1color) yscale(range(0 20 40 6
> 0 80 100)) ylabel(0 20 40 60 80 100) name(f4)

. *MENA
. twoway (scatter lagwe_mtech_av lagwe_lngdpc_av if year==2010 & region=="MENA", mlabel(is
> o3) mlabcolor(black) mlabposition(12) msymbol(i) mcolor(black) legend(off) xtitle(Lag GD
> P per capita) ytitle(Lag arms technology) title(MENA) ) (lfit lagwe_mtech lagwe_mtech  i
> f year==2010 & region=="MENA", lpattern(dash) lcolor(black) range(0 100) msymbol(iso3) m
> symbol(i) legend(off) ) , scheme(s1color) yscale(range(0 20 40 60 80 100)) ylabel(0 20 4
> 0 60 80 100) name(f5)

. *Africa
. twoway (scatter lagwe_mtech_av lagwe_lngdpc_av if year==2010 & region=="Sub-Saharan Afri
> ca", mlabel(iso3) mlabcolor(black) mlabposition(12) msymbol(i) mcolor(black) legend(off)
>  xtitle(Lag GDP per capita) ytitle(Lag arms technology) title(Sub-Saharan Africa) ) (lfi
> t lagwe_mtech lagwe_mtech  if year==2010 & region=="Sub-Saharan Africa", lpattern(dash) 
> lcolor(black) range(0 100) msymbol(iso3) msymbol(i) legend(off) ) , scheme(s1color) ysca
> le(range(0 20 40 60 80 100)) ylabel(0 20 40 60 80 100) name(f6)

. 
. graph combine f1 f2 f3 f4 f5 f6 ,  graphregion(color(white))

. 
. graph export "coercive_imbalance_figures/Figure3.eps", replace
file coercive_imbalance_figures/Figure3.eps saved as EPS format

. graph export "coercive_imbalance_figures/Figure3.pdf", replace
file coercive_imbalance_figures/Figure3.pdf saved as PDF format

. 
. graph drop *

. 
. *Calculate no of countries referenced in text (section 4)
. count if lagwe_mtech_av<lagwe_lngdpc_av & year==2010 & weu!=1 & independent=="yes"  
  129

. count if lagwe_mtech_av>lagwe_lngdpc_av & year==2010 & weu!=1 & independent=="yes" 
  21

. 
. // FIGURE 4: Average Adoption Lags for Arms and Civilian Technologies
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. 
. local categories "s a r t m h"

. local stechs "1 2 3 4 5 6 7 8"

. local atechs "1 2 3 4 5"

. local rtechs "1 2 3 4 5 6"

. local ttechs "1 2 3 4 5"

. local mtechs "1 2"

. local htechs "1 2"

. 
. //generate adoption lag if existed when tech invented
. foreach x of local categories {
  2. foreach y of local `x'techs {
  3. gen temp1=adoptionyear`x'`y'
  4. egen invented`x'`y'=min(temp1)
  5. gen temp2=1 if independent=="yes" & year==invented`x'`y'
  6. replace temp2=1 if independent=="yes" & year==invented`x'`y'+5
  7. egen existed_at_invetion=mean(temp2), by(country)
  8. gen adoptionlag`x'`y'=adoptionyear`x'`y'-invented`x'`y' if adoptionyear`x'`y'<3000  &
>  existed_at_invetion==1
  9. drop temp* existed_at_invetion
 10. }
 11. }
(25,940 missing values generated)
(33,389 missing values generated)
(0 real changes made)
(33,389 missing values generated)
(33,389 missing values generated)
(28,996 missing values generated)
(33,389 missing values generated)
(0 real changes made)
(33,389 missing values generated)
(33,389 missing values generated)
(27,850 missing values generated)
(33,389 missing values generated)
(0 real changes made)
(33,389 missing values generated)
(33,389 missing values generated)
(27,086 missing values generated)
(33,278 missing values generated)
(112 real changes made)
(11,773 missing values generated)
(27,086 missing values generated)
(25,749 missing values generated)
(33,278 missing values generated)
(109 real changes made)
(11,966 missing values generated)
(25,749 missing values generated)
(22,884 missing values generated)
(33,278 missing values generated)
(109 real changes made)
(11,966 missing values generated)
(22,884 missing values generated)
(24,470 missing values generated)
(33,279 missing values generated)
(109 real changes made)
(11,773 missing values generated)
(24,794 missing values generated)
(24,279 missing values generated)
(33,297 missing values generated)
(86 real changes made)
(15,633 missing values generated)
(24,279 missing values generated)
(30,524 missing values generated)
(33,389 missing values generated)
(0 real changes made)
(33,389 missing values generated)
(33,389 missing values generated)
(26,513 missing values generated)
(33,280 missing values generated)
(109 real changes made)
(12,159 missing values generated)
(26,513 missing values generated)
(27,659 missing values generated)
(33,280 missing values generated)
(109 real changes made)
(12,159 missing values generated)
(27,659 missing values generated)
(25,749 missing values generated)
(33,280 missing values generated)
(107 real changes made)
(12,352 missing values generated)
(25,749 missing values generated)
(26,189 missing values generated)
(33,312 missing values generated)
(72 real changes made)
(17,949 missing values generated)
(26,571 missing values generated)
(13,774 missing values generated)
(33,325 missing values generated)
(65 real changes made)
(20,458 missing values generated)
(21,987 missing values generated)
(14,729 missing values generated)
(33,317 missing values generated)
(69 real changes made)
(19,493 missing values generated)
(23,133 missing values generated)
(12,055 missing values generated)
(33,319 missing values generated)
(82 real changes made)
(17,563 missing values generated)
(19,886 missing values generated)
(12,819 missing values generated)
(33,305 missing values generated)
(91 real changes made)
(15,826 missing values generated)
(18,931 missing values generated)
(11,864 missing values generated)
(33,297 missing values generated)
(120 real changes made)
(10,229 missing values generated)
(15,493 missing values generated)
(18,931 missing values generated)
(33,250 missing values generated)
(148 real changes made)
(4,825 missing values generated)
(20,077 missing values generated)
(12,819 missing values generated)
(33,325 missing values generated)
(71 real changes made)
(19,107 missing values generated)
(21,414 missing values generated)
(12,628 missing values generated)
(33,320 missing values generated)
(73 real changes made)
(19,300 missing values generated)
(21,223 missing values generated)
(14,538 missing values generated)
(33,307 missing values generated)
(84 real changes made)
(17,177 missing values generated)
(21,414 missing values generated)
(20,077 missing values generated)
(33,276 missing values generated)
(125 real changes made)
(9,264 missing values generated)
(21,414 missing values generated)
(25,367 missing values generated)
(33,241 missing values generated)
(150 real changes made)
(4,439 missing values generated)
(25,749 missing values generated)
(26,762 missing values generated)
(33,279 missing values generated)
(109 real changes made)
(11,773 missing values generated)
(26,895 missing values generated)
(25,616 missing values generated)
(33,300 missing values generated)
(84 real changes made)
(16,019 missing values generated)
(25,616 missing values generated)
(18,549 missing values generated)
(33,259 missing values generated)
(139 real changes made)
(6,562 missing values generated)
(20,459 missing values generated)
(30,524 missing values generated)
(33,239 missing values generated)
(168 real changes made)
(965 missing values generated)
(30,524 missing values generated)

. 
. rename adoptionyears* adoptionyear1*

. rename adoptionyeara* adoptionyear2*

. rename adoptionyearr* adoptionyear3*

. rename adoptionyeart* adoptionyear4*

. rename adoptionyearm* adoptionyear5*

. rename adoptionyearh* adoptionyear6*

. 
. rename *lags* *lag1*

. rename *laga* *lag2*

. rename *lagr* *lag3*

. rename *lagt* *lag4*

. rename *lagm* *lag5*

. rename *lagh* *lag6*

. 
. rename inventeds* invented1*

. rename inventeda* invented2*

. rename inventedr* invented3*

. rename inventedt* invented4*

. rename inventedm* invented5*

. rename inventedh* invented6*

. 
. keep if year==2000
(33,216 observations deleted)

. keep adoptionlag* adoptionyear* invented*  country region

. 
. reshape long adoptionlag adoptionyear invented, i(country) j(tech)
(j = 11 12 13 14 15 16 17 18 19 21 22 23 24 25 31 32 33 34 35 36 41 42 43 44 45 51 52 61 6
> 2)
(variable adoptionlag19 not found)
(variable invented19 not found)

Data                               Wide   ->   Long
-----------------------------------------------------------------------------
Number of observations              173   ->   5,017       
Number of variables                  87   ->   6           
j variable (29 values)                    ->   tech
xij variables:
adoptionlag11 adoptionlag12 ... adoptionlag62->adoptionlag
adoptionyear11 adoptionyear12 ... adoptionyear62->adoptionyear
   invented11 invented12 ... invented62   ->   invented
-----------------------------------------------------------------------------

. 
. 
. 
. append using  "coercive_imbalance_raw\comin_adoption_lags.dta"
(variable tech was byte, now float to accommodate using data's values)

. 
. replace country="Bolivia, Plurinational State of" if country=="Bolivia"
(5 real changes made)

. replace country="Bosnia and Herzegovina" if country=="Bosnia-Herzegovina"
(1 real change made)

. replace country="Cote d'Ivoire" if country=="Cote D'Ivoire"
(29 real changes made)

. replace country="Czech Republic" if country=="Czechoslovakia"
(15 real changes made)

. replace country="Iran, Islamic Republic of" if country=="Iran"
(17 real changes made)

. replace country="Cote d'Ivoire" if country=="Ivory Coast"
(0 real changes made)

. replace country="Lao People's Democratic Republic" if country=="Laos"
(6 real changes made)

. replace country="Congo" if country=="Republic of the Congo"
(4 real changes made)

. replace country="Russian Federation" if country=="Russia"
(8 real changes made)

. replace country="Slovakia" if country=="Slovak Republic"
(3 real changes made)

. replace country="Korea, Republic of" if country=="South Korea"
(22 real changes made)

. replace country="Syrian Arab Republic" if country=="Syria"
(11 real changes made)

. replace country="Taiwan, Province of China" if country=="Taiwan"
(13 real changes made)

. replace country="Tanzania, United Republic of" if country=="Tanzania"
(6 real changes made)

. replace country="Venezuela, Bolivarian Republic of" if country=="Venezuala"
(15 real changes made)

. replace country="Viet nam" if country=="Vietnam"
(6 real changes made)

. 
. 
. ///create coercive adoption balanced to comin
> gen temp1=1 if tech>900 & adoptionlag!=. & invented<1910
(5,481 missing values generated)

. gen temp2=1 if tech>900 & adoptionlag!=. & invented>=1910 & invented<1940
(5,966 missing values generated)

. gen temp3=1 if tech>900 & adoptionlag!=. & invented>=1940
(5,906 missing values generated)

. egen sample1=sum(temp1), by(country)

. egen sample2=sum(temp2), by(country)

. egen sample3=sum(temp3), by(country)

. gen comin_sample=0

. replace comin_sample=1 if tech>900
(1,151 real changes made)

. replace comin_sample=1 if sample1>0 & invented<1910
(1,755 real changes made)

. replace comin_sample=1 if sample2>0 & invented>=1910 & invented<1940
(312 real changes made)

. replace comin_sample=1 if sample3>0 & invented>=1940
(902 real changes made)

. 
. gen comin_sample_adoptionlag=.
(6,168 missing values generated)

. replace comin_sample_adoptionlag=adoptionlag if comin_sample==1 & adoptionyear<=2002
(996 real changes made)

. 
. gen obs_adoption=1 if adoptionlag!=.
(3,778 missing values generated)

. 
. gen adoptionlag_nonwe_comin=adoptionlag if region!="Western Europe and off-shoots" & com
> in_sample==1  
(4,362 missing values generated)

. replace adoptionlag_nonwe_comin=adoptionlag if region!="Western Europe and off-shoots" &
>  comin_sample==1 & adoptionlag!=. 
(0 real changes made)

. gen coercive=1

. replace coercive=0 if tech>=900
(1,151 real changes made)

. 
. //Collapse
. collapse (mean) adoptionlag*  invented (sum) obs*, by(tech)

. 
. drop if tech==919 | tech==920 | tech==921 | tech==922 | tech==925 | tech==602 |  tech==4
> 05
(5 observations deleted)

. 
. drop if invented<1800 | invented==.
(7 observations deleted)

. gen military=1 if tech<900
(18 missing values generated)

. sort military invented

. gen temp=_n

. replace temp=temp-24 if military==.
(18 real changes made)

. tostring temp, force replace
temp was float now str2

. gen tech_fig="m" + temp if military==1
(18 missing values generated)

. replace tech_fig="c" + temp if military==.
(18 real changes made)

. 
. preserve

. 
. 
. set scheme plotplainblind 

. twoway scatter adoptionlag_nonwe_comin invented if tech<900 & invented>1810, mcolor(blac
> k) mlabel(tech_fig) mlabcolor(black) msymbol(circle) || ///
> scatter adoptionlag_nonwe_comin invented if tech>900 & invented>1810, mcolor(gs8) mlabel
> (tech_fig)  msymbol(circle) mlabcolor(gs8)  ///
> legend(label(1 "Military") label(2 "Civilian")  ) plotregion(fcolor(white)) graphregion(
> color(white)) ///
> xtitle("Year of first use anywhere") ytitle("Average adoption lag") ylabel(0(20) 80) ysc
> ale(range(0 80)) 

. 
. graph export "coercive_imbalance_figures/Figure4.eps", replace
file coercive_imbalance_figures/Figure4.eps saved as EPS format

. graph export "coercive_imbalance_figures/Figure4.pdf", replace
file coercive_imbalance_figures/Figure4.pdf saved as PDF format

. 
. restore

. 
. // FIGURE 5: Diffusion of Steam and Motor Ships for Military and Civilian Purposes
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_raw\ships.dta", clear

. 
. 
. destring ship*, replace force
ship_motor: contains nonnumeric characters; replaced as int
(41274 missing values generated)
ship_sail: contains nonnumeric characters; replaced as int
(39589 missing values generated)
ship_steam: contains nonnumeric characters; replaced as int
(40224 missing values generated)
ship_steammotor already numeric; no replace
shipton_motor already numeric; no replace
shipton_sail already numeric; no replace
shipton_steam already numeric; no replace
shipton_steammotor: contains nonnumeric characters; replaced as double
(38654 missing values generated)
ship_motor_navy already numeric; no replace
ship_sail_navy: contains nonnumeric characters; replaced as int
(41387 missing values generated)
ship_steam_navy: contains nonnumeric characters; replaced as int
(41379 missing values generated)

. drop if year<1800
(8,050 observations deleted)

. drop if year>1950
(9,338 observations deleted)

. 
. gen temp=year if ship_steam_navy!=.
(23,991 missing values generated)

. egen firstobs_steam=min(temp), by(country)
(15,553 missing values generated)

. drop temp

. gen temp=year if ship_sail_navy!=.
(23,999 missing values generated)

. egen firstobs_sail=min(temp), by(country)
(15,704 missing values generated)

. drop temp

. drop if firstobs_steam>1880
(19,328 observations deleted)

. 
. replace ship_motor=0 if year<1910
(3,630 real changes made)

. replace ship_motor_navy=0 if year<1910
(3,400 real changes made)

. replace ship_sail_navy=0  if ship_sail_navy+ship_steam_navy+ship_motor_navy==0 
(0 real changes made)

. replace ship_steam_navy=0  if ship_sail_navy+ship_steam_navy+ship_motor_navy==0 
(0 real changes made)

. replace ship_motor_navy=0  if ship_sail_navy+ship_steam_navy+ship_motor_navy==0 
(0 real changes made)

. replace ship_steammotor=ship_steam+ship_motor if ship_steam!=0 & ship_motor!=0
(323 real changes made, 154 to missing)

. replace ship_steammotor=ship_steam if ship_steammotor==.
(858 real changes made)

. replace ship_steammotor=ship_motor if ship_steammotor==.
(2,765 real changes made)

. replace shipton_steammotor=shipton_steam+shipton_motor if shipton_steam!=0 & shipton_mot
> or!=0
(906 real changes made, 731 to missing)

. replace shipton_steammotor=shipton_steam if shipton_steammotor==.
(920 real changes made)

. replace shipton_steammotor=shipton_motor if shipton_steammotor==.
(0 real changes made)

. 
. egen country_id=group(country)

. xtset country_id year

Panel variable: country_id (strongly balanced)
 Time variable: year, 1800 to 1950
         Delta: 1 unit

. 
. foreach x of varlist ship* {
  2. ipolate `x' year, by(country) gen(temp)
  3. replace `x'=temp
  4. drop temp
  5. }
(1078 missing values generated)
variable ship_motor was int now float
(106 real changes made)
(3158 missing values generated)
variable ship_sail was int now float
(346 real changes made)
(3526 missing values generated)
variable ship_steam was int now float
(417 real changes made)
(1009 missing values generated)
variable ship_steammotor was int now float
(62 real changes made)
(4797 missing values generated)
variable shipton_motor was int now float
(11 real changes made)
(3220 missing values generated)
(183 real changes made)
(3710 missing values generated)
(178 real changes made)
(3710 missing values generated)
(178 real changes made)
(1218 missing values generated)
variable ship_motor_navy was byte now float
(111 real changes made)
(3135 missing values generated)
variable ship_sail_navy was int now float
(1,596 real changes made)
(3107 missing values generated)
variable ship_steam_navy was int now float
(1,619 real changes made)

. 
. gen tradefleet=ship_sail+ship_steammotor
(3,495 missing values generated)

. gen tradeton=shipton_sail+shipton_steammotor
(3,800 missing values generated)

. gen navy=ship_sail_navy+ship_steam_navy+ship_motor_navy
(3,135 missing values generated)

. gen trade_steammotor=ship_steammotor/tradefleet
(3,495 missing values generated)

. gen tradeton_steammotor=shipton_steammotor/tradeton
(3,800 missing values generated)

. gen navy_sail=ship_sail_navy/navy
(3,381 missing values generated)

. gen navy_steammotor=1-navy_sail
(3,381 missing values generated)

. 
. replace navy_steammotor=0 if year<1820
(660 real changes made)

. replace navy_sail=1 if year<1820
(660 real changes made)

. replace navy_steammotor=1 if year>1900
(1,370 real changes made)

. replace navy_sail=0 if year>1900
(1,370 real changes made)

. ipolate navy_steammotor year, gen(temp) by(country)

. replace navy_steammotor=temp
(1,383 real changes made)

. drop temp

. ipolate navy_sail year, gen(temp) by(country)

. replace navy_sail=temp
(1,383 real changes made)

. drop temp

. 
. replace shipton_steammotor=. if ship_steammotor==.
(24 real changes made, 24 to missing)

. replace ship_steammotor=. if shipton_steammotor==.
(2,725 real changes made, 2,725 to missing)

. 
. collapse (mean) trade* navy* ship*, by(year)

. 
. gen tonnage_sail=shipton_sail/ship_sail

. gen tonnage_steam=shipton_steammotor/ship_steammotor
(15 missing values generated)

. gen navyton_sail=navy_sail*tonnage_sail/(navy_sail*tonnage_sail+navy_steammotor*tonnage_
> steam)
(15 missing values generated)

. gen navyton_steammotor=1-navyton_sail
(15 missing values generated)

. 
. twoway line navy_steammotor year, lcolor(black)  ///
> || line trade_steammotor year, lcolor(gs8)  ///
> || line navyton_steammotor year, lcolor(black)  lpattern(dash) ///
> || line tradeton_steammotor year, lcolor(gs8)  lpattern(dash) ///
> plotregion(fcolor(white)) graphregion(color(white)) ///
> xtitle("Year") ytitle("Share of ships or tonnage") ///
> legend(label(1 "Navy, % of ships")  label(2 "Merchant ships, % of ships")  label(3 "Navy
> , % of tonnage (est)") label(4 "Merchant ships, % of tonnage") pos(6)) scheme(s1color)

. 
. graph export "coercive_imbalance_figures/Figure5.eps", replace
file coercive_imbalance_figures/Figure5.eps saved as EPS format

. graph export "coercive_imbalance_figures/Figure5.pdf", replace
file coercive_imbalance_figures/Figure5.pdf saved as PDF format

. 
. // FIGURE 6: Western Europe and the Former Colonies
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. preserve

. keep if year>=1950 & year<=2010
(22,836 observations deleted)

. gen temp=year  if independent!="yes" & year<2010
(8,617 missing values generated)

. egen yearindp=max(temp), by(iso)
(5,124 missing values generated)

. replace yearindp=yearindp+1
(5,429 real changes made)

. gen ysi=year-yearindp
(5,124 missing values generated)

. drop temp

. 
. gen decade_indpendence=floor(yearindp/5)*5
(5,124 missing values generated)

. 
. keep if decade_indpendence<=1975 | decade_indpendence==.
(1,525 observations deleted)

. 
. drop if decade_indpendence==. & weu==0
(3,782 observations deleted)

. 
. egen tpopavg=mean(tpop), by(country)

. 
. drop if tpopavg<1000
(1,159 observations deleted)

. bysort yearindp: tab country

------------------------------------------------------------------------------------------
-> yearindp = 1954

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                               Cambodia |         61       50.00       50.00
                               Viet nam |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1955

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  Sudan |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1956

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                Morocco |         61       50.00       50.00
                                Tunisia |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1957

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  Ghana |         61       50.00       50.00
                               Malaysia |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1958

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                 Guinea |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1959

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  Niger |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1960

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  Benin |         61        5.56        5.56
                           Burkina Faso |         61        5.56       11.11
                                Burundi |         61        5.56       16.67
                               Cameroon |         61        5.56       22.22
               Central African Republic |         61        5.56       27.78
                                   Chad |         61        5.56       33.33
                                  Congo |         61        5.56       38.89
  Congo, the Democratic Republic of the |         61        5.56       44.44
                          Cote D'Ivoire |         61        5.56       50.00
                             Madagascar |         61        5.56       55.56
                                   Mali |         61        5.56       61.11
                             Mauritania |         61        5.56       66.67
                                Nigeria |         61        5.56       72.22
                                Senegal |         61        5.56       77.78
                              Singapore |         61        5.56       83.33
                                Somalia |         61        5.56       88.89
           Tanzania, United Republic of |         61        5.56       94.44
                                   Togo |         61        5.56      100.00
----------------------------------------+-----------------------------------
                                  Total |      1,098      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1961

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                           Sierra Leone |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1962

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                Algeria |         61       20.00       20.00
                                Jamaica |         61       20.00       40.00
                                 Rwanda |         61       20.00       60.00
                    Trinidad and Tobago |         61       20.00       80.00
                                 Uganda |         61       20.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        305      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1963

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  Kenya |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1964

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                 Malawi |         61       50.00       50.00
                                 Zambia |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1965

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                               Zimbabwe |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1966

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                               Botswana |         61       50.00       50.00
                                Lesotho |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1968

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                              Mauritius |         61       50.00       50.00
                                  Yemen |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1971

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                             Bangladesh |         61       50.00       50.00
                   United Arab Emirates |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1974

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                          Guinea-Bissau |         61      100.00      100.00
----------------------------------------+-----------------------------------
                                  Total |         61      100.00

------------------------------------------------------------------------------------------
-> yearindp = 1975

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                 Angola |         61       33.33       33.33
                             Mozambique |         61       33.33       66.67
                       Papua New Guinea |         61       33.33      100.00
----------------------------------------+-----------------------------------
                                  Total |        183      100.00

------------------------------------------------------------------------------------------
-> yearindp = .

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                              Australia |         61        5.00        5.00
                                Austria |         61        5.00       10.00
                                Belgium |         61        5.00       15.00
                                 Canada |         61        5.00       20.00
                                Denmark |         61        5.00       25.00
                                Finland |         61        5.00       30.00
                                 France |         61        5.00       35.00
                                Germany |         61        5.00       40.00
                                 Greece |         61        5.00       45.00
                                Ireland |         61        5.00       50.00
                                  Italy |         61        5.00       55.00
                            Netherlands |         61        5.00       60.00
                            New Zealand |         61        5.00       65.00
                                 Norway |         61        5.00       70.00
                               Portugal |         61        5.00       75.00
                                  Spain |         61        5.00       80.00
                                 Sweden |         61        5.00       85.00
                            Switzerland |         61        5.00       90.00
                         United Kingdom |         61        5.00       95.00
                          United States |         61        5.00      100.00
----------------------------------------+-----------------------------------
                                  Total |      1,220      100.00


. bysort decade_indpendence: tab country

------------------------------------------------------------------------------------------
-> decade_indpendence = 1950

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                               Cambodia |         61       50.00       50.00
                               Viet nam |         61       50.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        122      100.00

------------------------------------------------------------------------------------------
-> decade_indpendence = 1955

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                  Ghana |         61       14.29       14.29
                                 Guinea |         61       14.29       28.57
                               Malaysia |         61       14.29       42.86
                                Morocco |         61       14.29       57.14
                                  Niger |         61       14.29       71.43
                                  Sudan |         61       14.29       85.71
                                Tunisia |         61       14.29      100.00
----------------------------------------+-----------------------------------
                                  Total |        427      100.00

------------------------------------------------------------------------------------------
-> decade_indpendence = 1960

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                Algeria |         61        3.70        3.70
                                  Benin |         61        3.70        7.41
                           Burkina Faso |         61        3.70       11.11
                                Burundi |         61        3.70       14.81
                               Cameroon |         61        3.70       18.52
               Central African Republic |         61        3.70       22.22
                                   Chad |         61        3.70       25.93
                                  Congo |         61        3.70       29.63
  Congo, the Democratic Republic of the |         61        3.70       33.33
                          Cote D'Ivoire |         61        3.70       37.04
                                Jamaica |         61        3.70       40.74
                                  Kenya |         61        3.70       44.44
                             Madagascar |         61        3.70       48.15
                                 Malawi |         61        3.70       51.85
                                   Mali |         61        3.70       55.56
                             Mauritania |         61        3.70       59.26
                                Nigeria |         61        3.70       62.96
                                 Rwanda |         61        3.70       66.67
                                Senegal |         61        3.70       70.37
                           Sierra Leone |         61        3.70       74.07
                              Singapore |         61        3.70       77.78
                                Somalia |         61        3.70       81.48
           Tanzania, United Republic of |         61        3.70       85.19
                                   Togo |         61        3.70       88.89
                    Trinidad and Tobago |         61        3.70       92.59
                                 Uganda |         61        3.70       96.30
                                 Zambia |         61        3.70      100.00
----------------------------------------+-----------------------------------
                                  Total |      1,647      100.00

------------------------------------------------------------------------------------------
-> decade_indpendence = 1965

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                               Botswana |         61       20.00       20.00
                                Lesotho |         61       20.00       40.00
                              Mauritius |         61       20.00       60.00
                                  Yemen |         61       20.00       80.00
                               Zimbabwe |         61       20.00      100.00
----------------------------------------+-----------------------------------
                                  Total |        305      100.00

------------------------------------------------------------------------------------------
-> decade_indpendence = 1970

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                             Bangladesh |         61       33.33       33.33
                          Guinea-Bissau |         61       33.33       66.67
                   United Arab Emirates |         61       33.33      100.00
----------------------------------------+-----------------------------------
                                  Total |        183      100.00

------------------------------------------------------------------------------------------
-> decade_indpendence = 1975

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                                 Angola |         61       33.33       33.33
                             Mozambique |         61       33.33       66.67
                       Papua New Guinea |         61       33.33      100.00
----------------------------------------+-----------------------------------
                                  Total |        183      100.00

------------------------------------------------------------------------------------------
-> decade_indpendence = .

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                              Australia |         61        5.00        5.00
                                Austria |         61        5.00       10.00
                                Belgium |         61        5.00       15.00
                                 Canada |         61        5.00       20.00
                                Denmark |         61        5.00       25.00
                                Finland |         61        5.00       30.00
                                 France |         61        5.00       35.00
                                Germany |         61        5.00       40.00
                                 Greece |         61        5.00       45.00
                                Ireland |         61        5.00       50.00
                                  Italy |         61        5.00       55.00
                            Netherlands |         61        5.00       60.00
                            New Zealand |         61        5.00       65.00
                                 Norway |         61        5.00       70.00
                               Portugal |         61        5.00       75.00
                                  Spain |         61        5.00       80.00
                                 Sweden |         61        5.00       85.00
                            Switzerland |         61        5.00       90.00
                         United Kingdom |         61        5.00       95.00
                          United States |         61        5.00      100.00
----------------------------------------+-----------------------------------
                                  Total |      1,220      100.00


. 
. collapse (mean) mtech mtech_exf lagwe_mtech_exf lngdpc, by(decade_indpendence year)

. 
. replace mtech=. if decade_indpendence==1955 & year<1960
(5 real changes made, 5 to missing)

. replace mtech=. if decade_indpendence==1960 & year<1965
(5 real changes made, 5 to missing)

. replace mtech=. if decade_indpendence==1965 & year<1970
(5 real changes made, 5 to missing)

. replace mtech=. if decade_indpendence==1970 & year<1975
(4 real changes made, 4 to missing)

. replace mtech=. if decade_indpendence==1975 & year<1980
(5 real changes made, 5 to missing)

. 
. replace lngdpc=. if decade_indpendence==1955 & year<1960
(10 real changes made, 10 to missing)

. replace lngdpc=. if decade_indpendence==1960 & year<1965
(15 real changes made, 15 to missing)

. replace lngdpc=. if decade_indpendence==1965 & year<1970
(20 real changes made, 20 to missing)

. replace lngdpc=. if decade_indpendence==1970 & year<1975
(25 real changes made, 25 to missing)

. replace lngdpc=. if decade_indpendence==1975 & year<1980
(30 real changes made, 30 to missing)

. 
. drop if year<1960
(70 observations deleted)

. 
. twoway line mtech year if decade_indpendence==., lcolor(black) lpattern(solid)  || line 
> mtech year if decade_indpendence==1955, lcolor(gs6) lpattern(longdash) || line mtech yea
> r if decade_indpendence==1960, lcolor(gs6) lpattern(dash_dot) || line mtech year if deca
> de_indpendence==1965, lcolor(gs6) lpattern(dot) ///
> legend(label(1 "Western Europe") label(2 "Independent 1955-59") label(3 "Independent 196
> 0-64")  label(4 "Independent 1964-69") rows(2)) ///
> ytitle("Number of military techologies") xtitle("Year") plotregion(fcolor(white)) graphr
> egion(color(white)) title("Panel A: Military technology")  

. graph save Graph "panelA", replace
(file panelA.gph not found)
file panelA.gph saved

. 
. twoway line lngdpc year if decade_indpendence==., lcolor(black) lpattern(solid)  || line
>  lngdpc year if decade_indpendence==1955, lcolor(gs6) lpattern(longdash) || line lngdpc 
> year if decade_indpendence==1960, lcolor(gs6) lpattern(dash_dot) || line lngdpc year if 
> decade_indpendence==1965, lcolor(gs6) lpattern(dot) ///
> legend(label(1 "Western Europe") label(2 "Independent 1955-59") label(3 "Independent 196
> 0-64")  label(4 "Independent 1964-69") rows(2)) ///
> ytitle("Log GDP per capita") xtitle("Year") plotregion(fcolor(white)) graphregion(color(
> white)) title("Panel B: GDP per capita")  

. graph save Graph "panelB", replace
(file panelB.gph not found)
file panelB.gph saved

. 
. 
. grc1leg2 panelA.gph panelB.gph, iscale(1) altshrink rows(1) cols(2) legendfrom(panelA.gp
> h) graphregion(color(white))
-grc1leg2- working...

. graph export "coercive_imbalance_figures/Figure6.eps", replace
file coercive_imbalance_figures/Figure6.eps saved as EPS format

. graph export "coercive_imbalance_figures/Figure6.pdf", replace
file coercive_imbalance_figures/Figure6.pdf saved as PDF format

. erase panelA.gph 

. erase panelB.gph

. 
. restore

. 
. *###############################################################################
. *       5 ALL FIGURES IN THE ONLINE APPENDIX
. *###############################################################################
. 
. // Appendix FIGURE A1: How Much Are Other Countries Behind Western Europe in Arms Techno
> logy and Four Indicators of Modernization (balanced sample)
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. preserve

. drop if year<1850
(5,190 observations deleted)

. drop if year>2010
(346 observations deleted)

. 
. *gen balance non-we sample
. gen temp=1 if democracy!=.
(13,747 missing values generated)

. egen yrs_indep=sum(temp), by(iso3)

. drop temp

. 
. gen nonwe_bal=0

. replace nonwe_bal=1 if yrs_indep>160 & yrs_indep!=. & region!="Western Europe and off-sh
> oots"
(4,508 real changes made)

. 
. keep if nonwe_bal==1
(23,345 observations deleted)

. 
. tab country

                           Country name |      Freq.     Percent        Cum.
----------------------------------------+-----------------------------------
                            Afghanistan |        161        3.57        3.57
                              Argentina |        161        3.57        7.14
        Bolivia, Plurinational State of |        161        3.57       10.71
                                 Brazil |        161        3.57       14.29
                                  Chile |        161        3.57       17.86
                                  China |        161        3.57       21.43
                               Colombia |        161        3.57       25.00
                             Costa Rica |        161        3.57       28.57
                     Dominican Republic |        161        3.57       32.14
                                Ecuador |        161        3.57       35.71
                            El Salvador |        161        3.57       39.29
                              Guatemala |        161        3.57       42.86
                                  Haiti |        161        3.57       46.43
                               Honduras |        161        3.57       50.00
              Iran, Islamic Republic of |        161        3.57       53.57
                                  Japan |        161        3.57       57.14
                                Liberia |        161        3.57       60.71
                                 Mexico |        161        3.57       64.29
                                  Nepal |        161        3.57       67.86
                              Nicaragua |        161        3.57       71.43
                                   Oman |        161        3.57       75.00
                               Paraguay |        161        3.57       78.57
                                   Peru |        161        3.57       82.14
                     Russian Federation |        161        3.57       85.71
                               Thailand |        161        3.57       89.29
                                 Turkey |        161        3.57       92.86
                                Uruguay |        161        3.57       96.43
      Venezuela, Bolivarian Republic of |        161        3.57      100.00
----------------------------------------+-----------------------------------
                                  Total |      4,508      100.00

. 
. collapse (mean) lagwe_*, by(year)

. 
. sort year

. twoway line lagwe_mtech year , lcolor(black) lpattern(solid) || ///
> line lagwe_lngdpc year , lcolor(gs6) lpattern(solid) || ///
> line lagwe_naes year if lagwe_naes<=400, lcolor(gs6) lpattern(longdash)  || ///
> line lagwe_ups year , lcolor(gs6) lpattern(dash_dot) || ///
> line lagwe_schooling year ,lcolor(gs6) lpattern(dot)  ///
> legend(label(1 "Arms technology") label(2 "GDP per capita") label(3 "Industrialization")
>   label(4 "Urbanization") label(5 "Schooling")) ///
> ytitle("Lag to Western Europe (years)") xtitle("Year") plotregion(fcolor(white)) graphre
> gion(color(white)) scheme(s1color) yline(0)

. 
. graph export "coercive_imbalance_figures/FigureA1.eps", replace
file coercive_imbalance_figures/FigureA1.eps saved as EPS format

. graph export "coercive_imbalance_figures/FigureA1.pdf", replace
file coercive_imbalance_figures/FigureA1.pdf saved as PDF format

. 
. restore

. 
. 
. 
. // Appendix FIGURE A2: Modernization and Arms Technologies Across Regions
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. rename adoptionyears* adoptionyear1*

. rename adoptionyeara* adoptionyear2*

. rename adoptionyearr* adoptionyear3*

. rename adoptionyeart* adoptionyear4*

. rename adoptionyearm* adoptionyear5*

. rename adoptionyearh* adoptionyear6*

. 
. rename firstuses* firstuse1*

. rename firstusea* firstuse2*

. rename firstuser* firstuse3*

. rename firstuset* firstuse4*

. rename firstusem* firstuse5*

. rename firstuseh* firstuse6*

. 
. drop *_use

. drop if year<1700
(0 observations deleted)

. 
.  
. reshape long adoptionyear firstuse, i(year region* country naes ups lngdp iso3 schooling
> ) j(tech)
(j = 11 12 13 14 15 16 17 18 19 21 22 23 24 25 31 32 33 34 35 36 41 42 43 44 45 51 52 61 6
> 2)

Data                               Wide   ->   Long
-----------------------------------------------------------------------------
Number of observations           33,389   ->   968,281     
Number of variables                 161   ->   106         
j variable (29 values)                    ->   tech
xij variables:
adoptionyear11 adoptionyear12 ... adoptionyear62->adoptionyear
   firstuse11 firstuse12 ... firstuse62   ->   firstuse
-----------------------------------------------------------------------------

. egen invented=min(adoptionyear), by(tech) 

. 
. drop if invented<=1700
(133,556 observations deleted)

. 
. 
. *ordering
. gen temp=invented+tech/1000

. sort temp

. egen techorder=group(temp)

. drop temp

. sort country year techorder

. 
. ***Generate regional averages
. 
. *frontier -> update
. egen temp=mean(lngdp) if country=="France" | country=="United Kingdom" | country=="Germa
> ny", by(year)
(820,400 missing values generated)

. replace temp=. if invented!=year
(14,250 real changes made, 14,250 to missing)

. egen lngdp_adopt_frontier=mean(temp), by(tech)

. drop temp

. egen temp=mean(naes) if country=="France" | country=="United Kingdom" | country=="German
> y", by(year)
(820,400 missing values generated)

. replace temp=. if invented!=year
(14,250 real changes made, 14,250 to missing)

. egen naes_adopt_frontier=mean(temp), by(tech)

. drop temp

. egen temp=mean(ups) if country=="France" | country=="United Kingdom" | country=="Germany
> ", by(year)
(832,400 missing values generated)

. replace temp=. if invented!=year
(2,319 real changes made, 2,319 to missing)

. egen ups_adopt_frontier=mean(temp), by(tech)
(767,947 missing values generated)

. drop temp

. 
. *adoption
. egen adoption_region=mean(adoptionyear), by(tech region)
(54,255 missing values generated)

. gen temp=naes if year==adoptionyear 
(833,296 missing values generated)

. egen naes_adopt_region=mean(temp), by(region tech)
(106,780 missing values generated)

. replace temp=.
(1,429 real changes made, 1,429 to missing)

. replace temp=ups if year==adoptionyear  
(268 real changes made)

. egen ups_adopt_region=mean(temp), by(region tech)
(295,771 missing values generated)

. replace temp=.
(268 real changes made, 268 to missing)

. replace temp=lngdpc if year==adoptionyear  
(1,507 real changes made)

. egen lngdp_adopt_region=mean(temp), by(region tech)
(89,399 missing values generated)

. replace temp=.
(1,507 real changes made, 1,507 to missing)

. replace temp=schooling if year==adoptionyear  
(1,208 real changes made)

. egen edu_adopt_region=mean(temp), by(region tech)
(97,039 missing values generated)

. 
. sort techorder

. 
. *combined modernization figures, adoption (grc1leg2 needs to be installed)
. twoway line lngdp_adopt_region techorder if country=="France" & year==2000, lcolor(black
> ) lwidth(medthick) || ///
> line lngdp_adopt_region techorder if country=="Poland" & year==2000, lcolor(gs10) lwidth
> (medthick) || ///
> line lngdp_adopt_region techorder if country=="Turkey" & year==2000, lcolor(gs6) lwidth(
> medthick) || ///
> line lngdp_adopt_region techorder if country=="China" & year==2000, lcolor(gs6) lpattern
> (dash) lwidth(medthick) || ///
> line lngdp_adopt_region techorder if country=="Mali" & year==2000, lcolor(gs10) lpattern
> (dash) lwidth(medthick) || ///
> line lngdp_adopt_region techorder if country=="Brazil" & year==2000, lcolor(black) lpatt
> ern(dash) lwidth(medthick)  /// 
> legend(label(1 "W Europe and off-shoots") label(2 "Eastern Europe") label(3 "MENA") ///
> label(4 "Asia and Oceania") label(5 "Sub-Saharan Africa") label(6 "Latin America") ) ///
> plotregion(fcolor(white)) graphregion(color(white)) ///
> xtitle("Technology no." , size(small)) xlabel(, labsize(small)) ytitle("Log GDP per capi
> ta") ylabel(, labsize(small)) ///
> name(panela, replace) nodraw title("Panel A: GDP per capita", size(small) color(black) j
> ustification(center))

. 
. twoway line edu_adopt_region techorder if country=="France" & year==2000, lcolor(black) 
> lwidth(medthick) || ///
> line edu_adopt_region techorder if country=="Poland" & year==2000, lcolor(gs10) lwidth(m
> edthick) || ///
> line edu_adopt_region techorder if country=="Turkey" & year==2000, lcolor(gs6) lwidth(me
> dthick) || ///
> line edu_adopt_region techorder if country=="China" & year==2000, lcolor(gs6) lpattern(d
> ash) lwidth(medthick) || ///
> line edu_adopt_region techorder if country=="Mali" & year==2000, lcolor(gs10) lpattern(d
> ash) lwidth(medthick) || ///
> line edu_adopt_region techorder if country=="Brazil" & year==2000, lcolor(black) lpatter
> n(dash) lwidth(medthick)  /// 
> legend(label(1 "W Europe and off-shoots") label(2 "Eastern Europe") label(3 "MENA") ///
> label(4 "Asia and Oceania") label(5 "Sub-Saharan Africa") label(6 "Latin America") ) ///
> plotregion(fcolor(white)) graphregion(color(white)) ///
> xtitle("Technology no." , size(small)) xlabel(, labsize(small)) ytitle("Average years of
>  schooling") ylabel(, labsize(small)) ///
> name(panelb, replace) nodraw title("Panel B: Schooling", size(small) color(black) justif
> ication(center))

. 
. twoway line naes_adopt_region techorder if country=="France" & year==2000, lcolor(black)
>  lwidth(medthick) || ///
> line naes_adopt_region techorder if country=="Poland" & year==2000, lcolor(gs10) lwidth(
> medthick) || ///
> line naes_adopt_region techorder if country=="Turkey" & year==2000, lcolor(gs6) lwidth(m
> edthick) || ///
> line naes_adopt_region techorder if country=="China" & year==2000, lcolor(gs6) lpattern(
> dash) lwidth(medthick) || ///
> line naes_adopt_region techorder if country=="Mali" & year==2000, lcolor(gs10) lpattern(
> dash) lwidth(medthick) || ///
> line naes_adopt_region techorder if country=="Brazil" & year==2000, lcolor(black) lpatte
> rn(dash) lwidth(medthick)  /// 
> legend(label(1 "W Europe and off-shoots") label(2 "Eastern Europe") label(3 "MENA") ///
> label(4 "Asia and Oceania") label(5 "Sub-Saharan Africa") label(6 "Latin America") ) ///
> plotregion(fcolor(white)) graphregion(color(white)) ///
> xtitle("Technology no." , size(small)) xlabel(, labsize(small)) ytitle("Nonagricultural 
> employment") ylabel(, labsize(small)) ///
> name(panelc, replace) nodraw title("Panel C: Industrialization", size(small) color(black
> ) justification(center))

. 
. twoway line ups_adopt_region techorder if country=="France" & year==2000, lcolor(black) 
> lwidth(medthick) || ///
> line ups_adopt_region techorder if country=="Poland" & year==2000, lcolor(gs10) lwidth(m
> edthick) || ///
> line ups_adopt_region techorder if country=="Turkey" & year==2000, lcolor(gs6) lwidth(me
> dthick) || ///
> line ups_adopt_region techorder if country=="China" & year==2000, lcolor(gs6) lpattern(d
> ash) lwidth(medthick) || ///
> line ups_adopt_region techorder if country=="Mali" & year==2000, lcolor(gs10) lpattern(d
> ash) lwidth(medthick) || ///
> line ups_adopt_region techorder if country=="Brazil" & year==2000, lcolor(black) lpatter
> n(dash) lwidth(medthick)  /// 
> legend(label(1 "W Europe and off-shoots") label(2 "Eastern Europe") label(3 "MENA") ///
> label(4 "Asia and Oceania") label(5 "Sub-Saharan Africa") label(6 "Latin America") ) ///
> plotregion(fcolor(white)) graphregion(color(white)) ///
> xtitle("Technology no." , size(small)) xlabel(, labsize(small)) ytitle("Urbanization rat
> e") ylabel(, labsize(small)) ///
> name(paneld, replace) nodraw title("Panel D: Urbanization", size(small) color(black) jus
> tification(center))

. 
. grc1leg2 panela panelb panelc paneld, ///
> cols(2) plotregion(fcolor(white)) graphregion(color(white)) scheme(s1color) ///
> legendfrom(panela) lsize(small) 
-grc1leg2- working...

. 
. graph export "coercive_imbalance_figures/FigureA2.eps", replace
file coercive_imbalance_figures/FigureA2.eps saved as EPS format

. graph export "coercive_imbalance_figures/FigureA2.pdf", replace
file coercive_imbalance_figures/FigureA2.pdf saved as PDF format

. 
. // Appendix FIGURE A3: Arms Technology and Administrative Technology
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. by iso3 (year), sort: gen byte first_census = sum(census) == 1  & sum(census[_n - 1]) ==
>  0 

. by iso3 (year), sort: gen byte first_yearbook = sum( yearbook_published ) == 1  & sum( y
> earbook_published [_n - 1]) == 0

. egen firstyear_census = total(year * (first_census == 1)), by(iso3)

. egen firstyear_yearbook = total(year * (first_yearbook == 1)), by(iso3)

. drop census

. drop yearbook_published

. rename firstyear_census census

. replace census=. if census==0
(20,844 real changes made, 20,844 to missing)

. rename firstyear_yearbook yearbook

. replace yearbook=. if yearbook==0
(20,844 real changes made, 20,844 to missing)

. drop first_census

. drop first_yearbook

. 
. replace University=max(1300,University) if University<1300
(965 real changes made)

. 
. 
. gen ones=1

. egen region_no=group(region)
(404 missing values generated)

. 
. rename census Census

. rename yearbook Yearbook

. 
. replace region="W. Europe + off-shoots" if region=="Western Europe and off-shoots"
(4,393 real changes made)

. replace region="Eastern Europe" if region=="Eastern Europe and Balkan"
(3,571 real changes made)

. 
. foreach x of varlist Census Yearbook Diplo* University   { 
  2. preserve
  3. drop if year!=`x'
  4. collapse (sum) ones (mean) mtech region_no, by(region) 
  5. list region ones
  6. drop if ones<3
  7. graph bar mtech, horizontal over(region, sort(1) ) yscale(range(0 22)) ytitle("") ///
> plotregion(fcolor(white)) graphregion(color(white))  ylabel(, labsize(small)) ///
> name(`x', replace) title("`x'", color(black) justification(center)) legend(label(1 "Mili
> tary technologies in use"))
  8. restore
  9. 
. }
(33,324 observations deleted)

     +-------------------------------+
     |                 region   ones |
     |-------------------------------|
  1. |       Asia and Oceania     10 |
  2. |         Eastern Europe      5 |
  3. |          Latin America     19 |
  4. |                   MENA      9 |
  5. |     Sub-Saharan Africa      3 |
     |-------------------------------|
  6. | W. Europe + off-shoots     19 |
     +-------------------------------+
(0 observations deleted)
(33,324 observations deleted)

     +-------------------------------+
     |                 region   ones |
     |-------------------------------|
  1. |       Asia and Oceania     10 |
  2. |         Eastern Europe      5 |
  3. |          Latin America     19 |
  4. |                   MENA      9 |
  5. |     Sub-Saharan Africa      3 |
     |-------------------------------|
  6. | W. Europe + off-shoots     19 |
     +-------------------------------+
(0 observations deleted)
(33,226 observations deleted)

     +-------------------------------+
     |                 region   ones |
     |-------------------------------|
  1. |       Asia and Oceania     32 |
  2. |         Eastern Europe     17 |
  3. |          Latin America     27 |
  4. |                   MENA     23 |
  5. |     Sub-Saharan Africa     46 |
     |-------------------------------|
  6. | W. Europe + off-shoots     18 |
     +-------------------------------+
(0 observations deleted)
(33,304 observations deleted)

     +-------------------------------+
     |                 region   ones |
     |-------------------------------|
  1. |                             1 |
  2. |       Asia and Oceania     19 |
  3. |         Eastern Europe      8 |
  4. |          Latin America      9 |
  5. |                   MENA     17 |
     |-------------------------------|
  6. |     Sub-Saharan Africa     27 |
  7. | W. Europe + off-shoots      4 |
     +-------------------------------+
(1 observation deleted)

. 
. grc1leg2 Census Yearbook Diplomacy University, cols(2) plotregion(color(white)) graphreg
> ion(color(white)) xcommon ycommon 
-grc1leg2- working...

. 
. graph export "coercive_imbalance_figures/FigureA3.eps", replace
file coercive_imbalance_figures/FigureA3.eps saved as EPS format

. graph export "coercive_imbalance_figures/FigureA3.pdf", replace
file coercive_imbalance_figures/FigureA3.pdf saved as PDF format

. 
. *Calculate averages referenced in section 4 text
. encode region, gen(regionno)

. mean Census if Census!=. & year==2010, over(regionno)

Mean estimation                                        Number of obs = 65

-------------------------------------------------------------------------
                        |       Mean   Std. err.     [95% conf. interval]
------------------------+------------------------------------------------
      c.Census@regionno |
      Asia and Oceania  |     1918.3   9.988938      1898.345    1938.255
        Eastern Europe  |     1863.4   16.99588      1829.447    1897.353
         Latin America  |   1873.053   8.872546      1855.328    1890.778
                  MENA  |   1936.444   17.40938      1901.665    1971.224
    Sub-Saharan Africa  |   1973.667   6.385748       1960.91    1986.424
W. Europe + off-shoots  |   1835.053   4.972717      1825.118    1844.987
-------------------------------------------------------------------------

. mean Yearbook if Yearbook!=. & year==2010, over(regionno)

Mean estimation                                        Number of obs = 65

-------------------------------------------------------------------------
                        |       Mean   Std. err.     [95% conf. interval]
------------------------+------------------------------------------------
    c.Yearbook@regionno |
      Asia and Oceania  |     1930.2   14.54938      1901.134    1959.266
        Eastern Europe  |     1906.8   9.971961      1886.879    1926.721
         Latin America  |   1909.632   7.526254      1894.596    1924.667
                  MENA  |   1937.444   12.23736      1912.997    1961.891
    Sub-Saharan Africa  |       1971   9.539392      1951.943    1990.057
W. Europe + off-shoots  |   1873.263   5.879811      1861.517    1885.009
-------------------------------------------------------------------------

. 
. 
. // Appendix FIGURE A4: Examples of Weighted Adoption Rates and the Instrument
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_final\ci_main_jop.dta", clear

. 
. preserve

. gen tech_1=m2_iv_n2
(13,581 missing values generated)

. gen tech_2=r3_iv_n2
(13,135 missing values generated)

. 
. drop if year<1850 & year>2010
(0 observations deleted)

. 
. *Panel A
. twoway line tech_1 year if country=="Turkey" & year>1850 & year<2010, lcolor(black) ///
> || line tech_1 year if  country=="Brazil" & year>1850 & year<2010, lcolor(gs5)  ///
> || line tech_1 year if  country=="Thailand" & year>1850 & year<2010, lcolor(gs10) ///
> || line tech_2 year if country=="Turkey" & year>1850 & year<2010, lcolor(black)   ///
> || line tech_2 year if  country=="Brazil" & year>1850 & year<2010, lcolor(gs5)   ///
> || line tech_2 year if  country=="Thailand" & year>1850 & year<2010, lcolor(gs10)   ///
> plotregion(fcolor(white)) graphregion(color(white)) scheme(s1color) ///
> legend(order(1 "Turkey" 2 "Brazil" 3 "Thailand") rows(1) ) ///
> ytitle("Weighted adoption rate") xtitle("Year") title(Panel A: Weighted adoption rates) 
> ///
> text(0.9 1885  "Automatic" "machine gun") text(0.5 1980  "First generation" "jet fighter
> ")  name(panela, replace)

. 
. *Panel B
. twoway line mtech_exf year if country=="Turkey" & year>1850 & year<2010, lcolor(gs10) ||
>  line mtech_exf_iv_n2 year if country=="Turkey" & year>1850 & year<2010, lcolor(black) /
> //
> plotregion(fcolor(white)) graphregion(color(white)) scheme(s1color) ///
> legend(label(1 "Actual") label(2 "Instrument")  region(col(white))) ytitle("Military tec
> hnology index") xtitle("Year") name(panelb, replace) title(Panel B: Instrument Turkey)

. 
. resizecombine panela panelb, plotregion(fcolor(white)) graphregion(color(white)) xsize(1
> 2cm) ysize(16cm) cols(1)

. 
. graph export "coercive_imbalance_figures/FigureA4.eps", replace
file coercive_imbalance_figures/FigureA4.eps saved as EPS format

. graph export "coercive_imbalance_figures/FigureA4.pdf", replace
file coercive_imbalance_figures/FigureA4.pdf saved as PDF format

. 
. restore

. 
. 
. // Appendix FIGURE A5: Lower Cost per Effective Unit of Coercion
. cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance\Replication
C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication

. use "coercive_imbalance_raw\prices_jop.dta", clear

. 
. replace priceusdollar=price if currency=="US DOLLAR" & priceusdollar==.
(316 real changes made)

. drop if tech=="uncertain" | tech=="other" | tech==""
(119 observations deleted)

. 
. merge m:1 year using "coercive_imbalance_raw\cpi.dta"

    Result                      Number of obs
    -----------------------------------------
    Not matched                           175
        from master                         0  (_merge==1)
        from using                        175  (_merge==2)

    Matched                               404  (_merge==3)
    -----------------------------------------

. drop if _merge==2
(175 observations deleted)

. drop _merge

. 
. sort tech year

. egen techno=group(tech)

. 
. gen category=substr(tech,1,1)

. replace category="m" if tech=="f11" | tech=="f12"
(21 real changes made)

. replace category="h" if tech=="r08" | tech=="r09"
(12 real changes made)

. 
. gen originalprice=price

. 
. replace price=priceusdollar/usdgbp/uk_cpi*100
(404 real changes made, 29 to missing)

. replace price=pricepounds/uk_cpi*100 if price==.
(28 real changes made)

. 
. drop if price==.
(1 observation deleted)

. 
. drop if transactiontype=="Illicit"
(17 observations deleted)

. drop if transactiontype=="Used"
(5 observations deleted)

. drop if transactiontype=="Discounted"
(1 observation deleted)

. drop if techname=="uncertain"
(1 observation deleted)

. 
. 
. egen firstyear=min(year), by(tech)

. 
. egen meanprice=mean(price), by(tech)

. gen lnprice=ln(price)

. 
. gen bundle=0

. replace bundle=1 if transactiontype=="Bundled"
(14 real changes made)

. quietly: xi: reg price bundle i.tech if category=="a"

. gen price_unbundled=price-_b[bundle]*bundle

. 
. gen price_bundle=price if bundle==1
(365 missing values generated)

. gen price_nobundle=price if bundle==0
(14 missing values generated)

. 
. preserve 

. 
. local new = _N + 2

. set obs `new'
Number of observations (_N) was 379, now 381.

. replace techname="Matchlock musket" if _n==`new'-1
(1 real change made)

. replace tech="f03" if _n==`new'-1
(1 real change made)

. replace techname="Snaphaunce" if _n==`new'
(1 real change made)

. replace tech="f04" if _n==`new'
(1 real change made)

. replace category="f" if _n>`new'-2
(2 real changes made)

. replace buyer="US Government" if _n>`new'-2
(2 real changes made)

. sort tech year

. 
. 
. **Generate Panel A: Price of Small Arms
. graph hbar price if category=="f" & buyer=="US Government", over(techname, sort(firstyea
> r)) ///
> ytitle("Price (2010 USD)") plotregion(fcolor(white)) graphregion(color(white)) title(Pri
> ce of Small Arms) ///
> bar(1, color(gs8))   text(5 95 "N/a", place(e))  text(5 84 "N/a", place(e))

. graph save Graph "panelA", replace
(file panelA.gph not found)
file panelA.gph saved

. 
. **Generate Panel B: Price of Artillery
. graph hbar price if category=="a" & bundle==0, over(techname, sort(techno)  ) ///
> ytitle("Price (2010 USD)") plotregion(fcolor(white)) graphregion(color(white)) ///
> bar(1, color(gs8)) title(Price of Artillery)

. graph save Graph "panelB", replace
(file panelB.gph not found)
file panelB.gph saved

. restore

. 
. **Index construction firearms
. *cd "/Users/ngb815/Dropbox/Coercive_tech/stata/coercive_imbalance
. clear

. set obs 14
Number of observations (_N) was 0, now 14.

. 
. gen techname=""
(14 missing values generated)

. 
. replace techname="Matchlock musket" if _n==1
variable techname was str1 now str16
(1 real change made)

. replace techname="Snaphaunce" if _n==2
(1 real change made)

. replace techname="Flintlock" if _n==3
(1 real change made)

. replace techname="Percussion lock" if _n==4
(1 real change made)

. replace techname="Minié bullet rifle" if _n==5
variable techname was str16 now str19
(1 real change made)

. replace techname="Breechloading rifle" if _n==6
(1 real change made)

. replace techname="Tubular magazine rifle" if _n==7
variable techname was str19 now str22
(1 real change made)

. replace techname="Box magazine rifle" if _n==8
(1 real change made)

. replace techname="Assault rifle" if _n==9
(1 real change made)

. replace techname="Field gun" if _n==10
(1 real change made)

. replace techname="Rifled artillery" if _n==11
(1 real change made)

. replace techname="Steel tubes" if _n==12
(1 real change made)

. replace techname="Breechloading" if _n==13
(1 real change made)

. replace techname="Recoil mechanism" if _n==14
(1 real change made)

. 
. ** Effectiveness of Small Arms Technologies // Source here is: Zabecki, D.T., 2014. Germ
> any at War: 400 Years of Military History [4 volumes]. ABC-CLIO.
. *a. Range
. gen effective_range=50 if techname=="Matchlock musket"
(13 missing values generated)

. label variable effective_range "Effective firearms range in meters"

. replace effective_range=50 if techname=="Matchlock musket"
(0 real changes made)

. replace effective_range=50 if techname=="Snaphaunce"
(1 real change made)

. replace effective_range=75 if techname=="Flintlock"
(1 real change made)

. replace effective_range=75 if techname=="Percussion lock"
(1 real change made)

. replace effective_range=300 if techname=="Minié bullet rifle"
(1 real change made)

. replace effective_range=450 if techname=="Breechloading rifle"
(1 real change made)

. replace effective_range=500 if techname=="Tubular magazine rifle"
(1 real change made)

. replace effective_range=500 if techname=="Box magazine rifle"
(1 real change made)

. replace effective_range=400 if techname=="Assault rifle"
(1 real change made)

. *b. Rate of fire
. gen rate_of_fire=0.5 if techname=="Matchlock musket"
(13 missing values generated)

. label variable rate_of_fire "Max rate of fire firearms (shots/minute)"

. replace rate_of_fire=2 if techname=="Snaphaunce"
(1 real change made)

. replace rate_of_fire=3 if techname=="Flintlock" //I have subtracted 1 shot/minute to tak
> e misfire into account.
(1 real change made)

. replace rate_of_fire=4 if techname=="Percussion lock"
(1 real change made)

. replace rate_of_fire=4 if techname=="Minié bullet rifle"
(1 real change made)

. replace rate_of_fire=10 if techname=="Breechloading rifle"
(1 real change made)

. replace rate_of_fire=15 if techname=="Tubular magazine rifle"
(1 real change made)

. replace rate_of_fire=20 if techname=="Box magazine rifle" //changed to 20 from 17.5 -> 2
> 0 round trench magazines used in WWI, Before that I assume that 10 round magazine could 
> be fired and changed once
(1 real change made)

. replace rate_of_fire=90 if techname=="Assault rifle" //changed to 90 -> practical rate o
> f fire assuming magazine replaced twice in a minute. The rate of fire was otherwise list
> ed as 550 shots/min
(1 real change made)

. 
. **Index construction artillery
. *a. Range (shots per minute
. replace effective_range=(900+1480)/2 if techname=="Field gun" //source: https://en.wikip
> edia.org/wiki/Canon_de_12_Gribeauval AND https://en.wikipedia.org/wiki/Canon_obusier_de_
> 12
(1 real change made)

. replace effective_range=2800 if techname=="Rifled artillery" //source: https://en.wikipe
> dia.org/wiki/La_Hitte_system AND http://francoprussianwar.com/armies.htm
(1 real change made)

. replace effective_range=3800 if techname=="Steel tubes" //source: Source: Bruce I- Gudmu
> ndsson "On Artillery"https://www.questia.com/read/28495535/on-artillery
(1 real change made)

. replace effective_range=(3000+3000+5000)/3 if techname=="Breechloading"
(1 real change made)

. replace effective_range=(7500+8500)/2 if techname=="Recoil mechanism" // Sources: https:
> //en.wikipedia.org/wiki/76_mm_gun_M1900 + https://en.wikipedia.org/wiki/Canon_de_75_mod%
> C3%A8le_1897 
(1 real change made)

. *b. Rate of fire
. replace rate_of_fire=1 if techname=="Field gun"
(1 real change made)

. replace rate_of_fire= 1 if techname=="Rifled artillery"
(1 real change made)

. replace rate_of_fire=2 if techname=="Steel tubes" // sources say 2-3. Conservatively set
>  to 2.
(1 real change made)

. replace rate_of_fire=2 if  techname=="Breechloading" // https://en.wikipedia.org/wiki/De
> _Bange_90_mm_cannon
(1 real change made)

. replace rate_of_fire=8 if techname=="Recoil mechanism" //average of sustained and burst.
>  Could in theory fire three times as often
(1 real change made)

. 
. gen simpleeffect=sqrt(effective_range*rate_of_fire)

. 
. gen category=""
(14 missing values generated)

. replace category="f" if _n<10
(9 real changes made)

. replace category="a" if _n>9
(5 real changes made)

. 
. graph hbar simpleeffect if category=="f", over(techname, sort(simpleeffect)) ///
> ytitle("Geometric mean of range and rate of fire") plotregion(fcolor(white)) graphregion
> (color(white)) ///
> bar(1, color(gs8)) title(Effectiveness of Small Arms)

. graph save Graph "panelC", replace
(file panelC.gph not found)
file panelC.gph saved

. 
. graph hbar simpleeffect if category=="a", over(techname, sort(simpleeffect)) ///
> ytitle("Geometric mean of range and rate of fire") plotregion(fcolor(white)) graphregion
> (color(white)) ///
> bar(1, color(gs8)) title(Effectiveness of Artillery)

. graph save Graph "panelD", replace
(file panelD.gph not found)
file panelD.gph saved

. 
. graph combine "panelA.gph" "panelB.gph" "panelC.gph" "panelD.gph", plotregion(fcolor(whi
> te)) graphregion(color(white))

. 
. graph export "coercive_imbalance_figures/FigureA5.eps", replace
file coercive_imbalance_figures/FigureA5.eps saved as EPS format

. graph export "coercive_imbalance_figures/FigureA5.pdf", replace
file coercive_imbalance_figures/FigureA5.pdf saved as PDF format

. 
. 
. erase panelA.gph 

. erase panelB.gph

. erase panelC.gph 

. erase panelD.gph

. 
. log off
      name:  <unnamed>
       log:  C:\Users\ngb815\Dropbox\Coercive_tech\stata\coercive_imbalance\Replication\re
> sults_coercive_imbalance.log
  log type:  text
 paused on:  17 Aug 2023, 12:28:07
------------------------------------------------------------------------------------------
